Multi-hydrolase knockout

Abstract

The present invention relates to a combined multi-hydrolase knock-out and gene minimization approach to improve CHO host cells for recombinant protein production. This approach combines genomic excision with removal of phenotypes that are disadvantageous for bioprocessing, such as polysorbate degrading hydrolases.

Description

[0001] Multi-hydrolase knockout

[0002] FIELD OF THE INVENTION

[0003]

[0001] The present invention relates to a combined multi-hydrolase knockout and gene minimization approach to improve CHO host cells for recombinant protein production. This approach combines genomic excision with removal of phenotypes that are disadvantageous for bioprocessing, such as polysorbate degrading hydrolases.

[0004] BACKGROUND

[0005]

[0002] In pharmaceutical bioprocesses host cell proteins (HCPs) have been challenging. Originating from the manufacturing cell line, most commonly the Chinese hamster ovary (CHO) cell line, HCPs may cause roadblocks at several steps during the biopharmaceutical production. Especially immunogenic or enzymatically active HCPs which are difficult to remove during the downstream process can compromise product stability, product integrity and / or product safety. In recent years, enzymatic degradation of polysorbate (PS) by hydrolytic HCPs has been observed in several drug products across the biopharmaceutical landscape (Roy, I. et al., J Pharm Sci, 2021 , 110: 3313-3323). Enzymatic polysorbate (PS) degradation in biological drug formulations mediated by host cell proteins (HCPs) has therefore become a widespread and severe challenge in the biopharmaceutical industry.

[0006]

[0003] As an indispensable surfactant, PS is used in drug formulations to prevent biopharmaceutical agents - often monoclonal antibodies or more recently antibody-like formats - from aggregation and minimizes surface adhesion and air-liquid interfacial stress. Hydrolytic HCPs, known as hydrolases, hydrolyze the ester bond in PS canceling out the protective effect of PS and releasing free fatty acids, which can aggregate and form particles in the final drug product. Unfortunately, even trace amounts of hydrolases are acting on PS over time rendering the identification of responsible culprit hydrolases a great challenge.

[0007]

[0004] PS is a heterogeneous mixture consisting of a hydrophilic polyoxyethylene sorbitan head group that is esterified to one or multiple hydrophobic fatty acids of varying length. Degradation of PS is driven via two independent pathways: Chemical degradation by autooxidation or hydrolysis and enzymatic degradation elicited by hydrolytic enzymes, which are mostly secreted during the bioprocess. Whereas chemical degradation has been successfully prevented by the use of buffering agents and antioxidants, hydrolytic enzymes acting on ester bonds (Enzyme class 3.1) turned out to be the major root cause of PS degradation. (Dixit N et al., J Pharm Sci 2016; 105: 1657-66; Hall T et al., J Pharm Sci 2016; 105: 1633-42; McShan AC et al., J Pharm Sci Technol 2016; 70: 332-45; Zhang S et al., J Pharm Sci 2020; 109: 3300-07).

[0008]

[0005] In recent years, significant effort has been put into the identification and characterization of potential HCPs acting on ester bonds of PS (Graf T. et al. J Pharm Sci 2021 ; 110: 3558-67; Madsen JA, mAbs 2015; 7: 1128-37; Mbrtstedt H et al., Journal of Pharmaceutical and Biomedical Analysis 2020; 185: 113256; Thompson JH et al., Rapid Communications in Mass Spectrometry 2014; 28: 855-60; Li X et al. Analytical Chemistry 2021 ; 93: 8161-69; Liu G-Y et al., Analytical Chemistry 2022; 94: 8625-32). However, the verification that a hydrolase is involved in PS degradation is difficult. One prominent example is phospholipase B-Like 2 (PLBL2), which was long considered an active polysorbate degrading HCP. However, contaminated recombinant HCP samples resulted in misinterpretations, emphasizing the importance of a well-designed purification and characterization strategy for recombinant HCPs to avoid misleading data interpretation as a result of persisting impurities. (Zhang S et al., J Pharm Sci 2020; 109: 2710-18).

[0009]

[0006] As hydrolytic HCPs were found to cause PS degradation at trace levels, <10 ppm (Zhang S et al., J Pharm Sci 2020; 109: 3300-07) detection and characterization of these low abundant enzymes represents a major analytical challenge for the biopharmaceutical community. Therefore, new analytical methods were developed including optimized sample preparation prior to mass spectrometric analysis and activity- based- protein profiling.

[0010]

[0007] To date, advances in HCP detection resulted in a more comprehensive list of potentially PS-degrading protein contaminants detected at various stages of the bioprocess. Yet, data regarding their enzymatic activity profiles against PS and an understanding of the impact of the downstream purification processes on these enzymes is often still lacking. A less explored but promising approach to address this issue is the recombinant overexpression and simultaneous screening of multiple putative polysorbate-degrading HCPs, especially the hydrolase species.

[0011]

[0008] Two main mechanisms are rendering HCPs “difficult to remove”: product association or product co-elution (Singh, S.K. et al., Biotechnology progress, 2020, 36, e2936). Stringent wash buffers including extreme pH values or wash additives are commonly applied to mitigate these retention issues (Chollangi, S. et al., Biotechnology and bioengineering, 2015, 112: 2292-2304; Li, Y., Protein expression and purification, 2017, 134: 96-103; Shukla, A. A. and Hinckley, P., Biotechnology progress, 2008, 24: 1115-1121). However, it was shown that the inherent properties of monoclonal antibodies (mAbs) require a tailored purification process which, even when highly optimized, is still unlikely to completely deplete all conceivable HCPs (Maier, M. et al., Biotechnology and bioengineering, 2024, 121 (10): 3181- 3195; Oh, Y.H. et al., Biotechnology and bioengineering, 2024, 121(4), 1284-1297). Traceless and simultaneous removal of several hydrolytic HCPs is even more challenging for the more diverse antibody-like molecules, including bispecific and / or single-chain formats, which may need adapted purification trains due to their diverse physiochemical properties.

[0012]

[0009] An additional tackling strategy could be applied during upstream processing, where cell viabilities were reported to have the largest influence on the HCP profile as HCPs can be intracellular proteins released by apoptotic / dead cells. Thus, by varying the process format including parameters like culture media, the feeding regime, culture duration and / or the harvest strategy HCP abundance in the harvest could be reduced. Another possible mitigation strategy is the identification and removal of such critical HCPs in the production cell line.

[0013]

[0010] Engineering of mammalian cell lines and especially CHO cells has a long history and CHO cells have been altered to obtain and / or lose a certain phenotype or characteristics by overexpression, downregulation or knockout (KO) of individual genes or combinations thereof, such as to improve productivity, controlling product quality, such as glycosylation and ensuring cell line stability. Although knockout techniques are known in the art, multi-gene knockouts are still not trivial and bear a high risk of negatively affecting cell culture parameters, such as viability and / or productivity of the cell, and hence the suitability of the cell line as production cell line. This risk is even higher for multiple genes encoding related proteins, particularly if proteins with related and potentially redundant function are deleted. To date, there is only one report on KO of a hydrolytic HCP-lipoprotein lipase (LPL)- to reduce PS degradation (Chiu, J. et al., Biotech & Bioengineering, 2017, 114: 1006-1015). Yet, for a production cell line suitable as platform production cell line multi-gene knockouts of hydrolytic HCP are required. Only few attempts have been made to combine the KO of multiple HCPs of which only some reports described a KO of more than one hydrolases (WO 2921 / 195464 A2; Carver et al., Biotechnology progress 38(1), e3212, doi:10.1002 / btpr.3212) and without investigating the effect of the KO cell line on PS degradation and / or product quality.

[0014]

[0011] At the same time, reduction of the complexity of CHO genomes would be desirable as CHO cells have evolved as part of a multicellular organism and thus, harbor many cellular functions irrelevant (i.e. , not essential) for their application as production hosts in industrial bioprocesses. These unnecessary genes pose a significant metabolic demand resulting in reduced growth, cell density and / or productivity. However, the largest reported genomic excision in CHO cells so far has only 864 kb (Jerabek, T. et al., New Biotechnology, 2024, 79: 100-110).

[0015] Thus, there is also a need for the identification of locations for large genomic excisions that can contribute to an overall smaller genome to generate a “leaner” CHO cell line, which is ideally combined with the removal of HCPs, particularly hydrolases, that negatively impact cell performance towards a production cell line with overall superior bioprocess characteristics.

[0016] SUMMARY OF THE INVENTION

[0017]

[0012] The aim of the present invention is to provide a cell line engineering approach for simultaneously knocking-out multiple hydrolases acting on ester bonds, such as lipases, in mammalian cells, particularly CHO cells, to reduce hydrolase activity, particularly PS degrading activity, and combining this with a genome minimization approach by excising entire gene clusters containing dispensable and undesired genes. In the present invention up to nine critical hydrolases were knocked-out in a CHO production cell line, including entire carboxylesterase cluster excisions, exhibiting a total genomic size of > 1 megabases. Surprisingly this multi-hydrolase knockout CHO cell line, which lacked in total 50 genes in comparison to the parental strain, is able to deliver high antibody titers without compromising product quality. Importantly, hydrolytic activity and thus PS degradation could be dramatically reduced. This allows a universal and product-independent solution for the elimination of PS- degrading hydrolases in a CHO-based bioprocess.

[0018]

[0013] The present invention relates to a CHO cell comprising, (i) a lipoprotein lipase (LPL) gene knockout (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) of the CHO genome, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). Thus, two clusters each comprising carboxylesterase genes are deleted in said CHO cell. In certain embodiments at least one of the two or more genome excisions removes 1000 kb or more, preferably 1500 kb or more of the CHO genome. Preferably, the LPL knockout, the at least three hydrolase knockouts and the two or more genome excisions are biallelic.

[0019]

[0014] In certain embodiments the two or more genome excisions are genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). The two or more genome excisions are genome excisions (a) of the CES1 cluster that remove at least 1000 kb and (b) of the CES2 cluster that remove at least 500 kb of the CHO genome. The two or more genome excisions deleting at least two gene clusters comprising a carboxylesterase gene may additionally comprise upstream and / or downstream flanking genomic regions, wherein the upstream and / or downstream flanking genomic regions are preferably 200 kb, 300 kb or 600 kb or more. The CHO cell may further comprise at least one additional genome excision of a gene cluster and / or of at least 800 kb of the CHO cell genome.

[0020]

[0015] In certain embodiments the CHO cell further comprises a glutamine synthetase (GS) knockout. The gene knockouts in the CHO cell according to the invention may be gene excision and / or gene inactivation.

[0021]

[0016] In the CHO cell according to the invention, in certain embodiments the (ii) at least three further hydrolase gene knockouts comprise at least three knockouts of genes coding for hydrolases (a) selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2), and / or (b) of the hydrolase class with the EC number 3.1 and comprising a catalytical triad consisting of a serine (Ser)-histidine (His)-aspartate / glutamate (Asp / Glu) motif. In more specific embodiments the (ii) at least three further hydrolase gene knockouts comprise knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate- hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2). In even more specific embodiments, the (ii) at least three further hydrolase gene knockouts comprise knockouts of hydrolase genes selected from the group consisting of lipase A (LIPA), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2).

[0022]

[0017] In another embodiment, the CHO cell comprises (i) a lipoprotein lipase (LPL) gene knockout, (ii) at least five further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene. In more specific embodiments the (ii) at least five further hydrolase gene knockouts comprise at least three knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH 1 ) , lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2). In even more specific embodiments, the (ii) at least five further hydrolase gene knockouts comprise knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2).

[0023]

[0018] The CHO according to the invention may further comprise a gene knockout of the gene coding for alpha-(1 ,6)-fucosyltransferase (FUT8) and / or a glutamine synthetase gene knockout. In certain embodiments the CHO cell may also comprise (optionally in addition to the FUT8 and / or GS knockout) a Bax and Bak1 gene knockout. In certain embodiments the CHO cell according to the invention further comprises a recombinant gene encoding a therapeutic protein.

[0024]

[0019] In another aspect, the invention relates to a method of producing a recombinant protein, comprising, (a) introducing a nucleic acid comprising a gene coding for a therapeutic protein and a gene coding for a selection marker into the CHO cell of the invention; (b) culturing the cell of step (a) under conditions to produce the therapeutic protein; (c) harvesting the therapeutic protein; and (d) purifying and optionally formulating the therapeutic protein.

[0025]

[0020] In yet another aspect the invention relates to a pharmaceutical composition comprising a recombinant protein produced according to the method of the invention.

[0026]

[0021] In yet another aspect, the invention relates to a method of reducing polysorbate degrading activity in a CHO cell, comprising (a) providing a CHO cell; (b) knocking out (i) lipoprotein lipase (LPL) gene and (ii) at least three further hydrolase gene knockouts in the CHO cell; and (c) excising genomic regions comprising (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) in the CHO cell, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C).

[0027]

[0022] In yet another aspect, the invention provides a method of reducing polysorbate degrading activity in a recombinant product produced in a CHO cell, wherein the method comprises (a) providing a CHO cell; (b) knocking out (i) lipoprotein lipase (LPL) gene and (ii) at least three further hydrolase gene knockouts in the CHO cell; (c) excising genomic regions comprising (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) in the CHO cell, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C); (d) transfecting the CHO cell comprising the knockouts of step (b) and the two or more genome excisions of step (c) with a gene coding for a therapeutic protein and a gene coding for a selection marker; (e) harvesting the therapeutic protein; and (f) purifying and optionally formulating the therapeutic protein.

[0028]

[0023] In certain embodiments the method uses the CHO cell according to the invention.

[0024] In yet another embodiment the invention relates to the use of the CHO cell according to the invention for use in producing a recombinant protein.

[0029] DESCRIPTION OF THE FIGURES

[0030]

[0025] Figure 1 : (A) Gene expression data in transcripts per million of potentially PS- degrading hydrolases identified in drug products. Transcript data of producer cell lines from three different campaigns are depicted. One clone each (two timepoints, n=2) was analyzed for monoclonal antibody (mAb) 1 and 2 and a producer pool (one timepoint, n=24) was analyzed for mAb 3. Data are sorted in descending order according to their occurrence in 7 investigated drug products. (B) Relative protein abundance (MS1 intensity) of potentially PS- degrading hydrolases identified in drug products. Protein expression data in cell culture supernatant from three different campaigns are depicted. One clone each (technical triplicates, n=3) was analyzed for mAb 1 and 2 and a producer pool (technical duplicates, n=2) was analyzed for mAb 3.

[0031]

[0026] Figure 2: (A) Agarose gel electrophoresis for verification of biallelic lipoprotein lipase (LPL) knockout (KO) in mAb producer LPL KO (KO) versus mAb producer LPL wildtype (wt). Deletion PCR and non-deletion PCR were performed with genomic DNA. Lack of bands in non-deletion PCRs indicates biallelic deletion. M = Marker. (B) LPL concentration in day 3 fed-batch samples of mAb producer cell lines with and without LPL KO. Cell lines were cultured head-to-head in an ambr®250 system and harvested cell culture supernatant was subjected to an LPL ELISA. (C) Polysorbate degradation in fed-batch samples of mAb producer cell lines with (left columns) and without (right columns) LPL assayed over a 14 days time frame. Cell lines were cultured head-to-head over 14 days in an ambr®250 system and harvested cell culture supernatant was subjected to a fluorescence micelle assay. Statistical significance was tested using ordinary two-way ANOVA with Sidak's Multiple Comparison Test, p < 0.05 was considered as statistically significant, ns = not significant.

[0032]

[0027] Figure 3: (A) LPL concentration in day 3 fed-batch samples of mAb host cell lines with and without LPL KO. Cell lines were cultured head-to-head in an ambr®15 system and harvested cell culture supernatant was subjected to an LPL ELISA. (B) Hydrolase expression plasmid suitable for transposase-mediated stable transfection via inverted repeats recognition sites. His- and Strep- tags were added at the 3‘-end of the sequence coding for the native hydrolase. Glutamine synthetase was used for stable selection. Created by BioRender. (C) Relative protein amounts of expressed hydrolases as the ratio of intracellular (hatched bar) to extracellular (white bar) protein. Hydrolases were quantified using biolayer interferometry with his2 sensors on the Octet system. Average values of two replicates with standard deviations are shown.

[0028] Figure 4: (A) Polysorbate (PS) degradation over time of crude cell culture supernatant of hydrolase expressing cell lines determined via the fluorescence micelle assay. Cell culture supernatants of (fed-)batch fermentations were diluted in PS-spiked fresh cultivation medium and degradation was monitored over 14 days (dO, d1 , d3, d7, d14). PS concentration was normalized to the initial PS concentration on dO. Average values of four technical replicates with standard deviations are shown. (B) PS degradation over time of purified hydrolases determined via the fluorescence micelle assay. Purified hydrolases were spiked into a formulation buffer to a final concentration of 20 pg / mL and degradation was monitored over 14 days (dO, d1 , d3, d7, d14). PS concentration was normalized to the initial PS concentration on dO. Average values of four technical replicates with standard deviations are shown.

[0033]

[0029] Figure 5: (A) pH dependent activity of the eight active hydrolases after purification. Values are normalized to highest activity of the respective hydrolase. Average values of duplicate analyses are shown. (B) Normalized hydrolytic activities (MLID4 Assay) of three mAb samples after the second polishing step at different pH values. Values are normalized to highest activity of the respective mAb sample. Average values of duplicate analyses are shown.

[0034]

[0030] Figure 6: Knockout (KO) strategy for the generation of a CHO K1 9x hydrolase KO cell line. Based on a CHO K1 cell line lipoprotein lipase (LPL) was knocked out using zinc finger nucleases, inducing insertion and / or deletions (InDei) in the target gene. Carboxylesterase (Ces) variants were knocked out via a genome cluster excision using 2x sgRNAs combined with Cas9. Platelet-activating factor acetylhydrolase (Pla2g7), Phospholipase A2 group XV (Pla2g15), Lipase A (Lipa), Isoamyl acetate hydrolyzing esterase 1 (Iah1) and Palmitoyl- protein thioesterase 1 (Ppt1) were knocked out via a single sgRNA / Cas9 complex, inducing InDeis. The two genes Pla2g15 / Lipa and Iah1 / Ppt1 were each knocked out simultaneously. The RNA sequence within Figure 6 is provided as SEQ ID NO: 23.

[0035]

[0031] Figure 7: Hydrolase gene expression analysis of the starting clone, intermediate clones (A) and final 9x KO clones (B) of the multi hydrolase KO process. Successive clones are separated by a dotted line and with each new intermedia clone gene expression of the respectively knocked out hydrolases is reduced. Clones were measured once or in biological duplicates of independent ambr run samples (d4), indicated by standard deviations, run samples (d4), indicated by standard deviations.

[0036]

[0032] Figure 8: Ambr15® fed-batch cultivation of a selection of hydrolase knockout (KO) cell lines in comparison to the parental CHO K1 host cell line. (A) Viable cell density (VCD) and (B) viability over 14 days fed-batch. Each point represents a daily measurement. Day 2 and 3 were not measured. Biological duplicates were cultivated and are represented as two separate lines per clone.

[0033] Figure 9: Fluorescence micelle assay (FMA) with harvested cell culture fluid (HCCF) samples from an ambr cultivation. HCCF from the clones as indicated (see Table 2) were diluted in fresh cultivation medium spiked with polysorbate (PS) and incubated for 0, 1 , 3, 7 and 14 days at room temperature, represented by 5 bars per sample in the diagram from left to right. For the 9xKO cells (Multi KO) HCCF from 6 different clones (1 D8, 1 F8, 1G1 , 2B4, 2C3 and 2G11) were analysed. Polysorbate degradation over time was measured in 4 technical replicates. Values were normalized to starting concentration on dO and are depicted as mean ± standard deviation.

[0037]

[0034] Figure 10: Fed-batch data from lgG1 producing 9x hydrolase KO cell pools. Parental host K1 clone and six 9x hydrolase KO clones were stably transfected with lgG1 coding plasmids and resulting producer pools were compared in a fed-batch run. Viable cell density (VCD) (A) and viability (B) were measured from day 3 until harvest on day 11 . Volumetric productivity (C) and specific productivity (qp) (D) were determined from day 6 to day 11 . Cell pools were cultivated in biological duplicates (n=2) and data is shown as mean ± standard deviation.

[0038]

[0035] Figure 11 : Fed-batch data from lgG4 producing 9x hydrolase KO cell pools. Parental host K1 clone and six 9x hydrolase KO clones were stably transfected with lgG4 coding plasmids and resulting producer pools were compared in a fed-batch run. Viable cell density (VCD) (A) and viability (B) were measured from day 3 until harvest on day 11 . Volumetric productivity (C) and specific productivity (D) were determined from day 6 to day 11. Cell pools were cultivated in biological duplicates (n=2) and data is shown as mean ± standard deviation.

[0039]

[0036] Figure 12: Host cell protein (HCP) quantification in harvested cell culture fluid (HCCF) (A) and protein A purified monoclonal antibody (mAb) fraction (B). IgG 1 and lgG4 producing 9x hydrolase KO and parental CHO K1 cell pools were fed-batch cultivated and HCCF was assayed directly and after protein A purification. Biological duplicates (n=2) were measured via an HCP Octet® assay and are depicted as mean ± standard deviation.

[0040]

[0037] Figure 13: 4-methylumbelliferyl decanoate (MLID4) assay in harvested cell culture fluid (HCCF) at pH 7.4 (A) and protein A purified monoclonal antibody (mAb) fraction (B). lgG1 and lgG4 producing 9x hydrolase KO and parental CHO K1 cell pools were fed-batch cultivated and HCCF was assayed directly and after protein A purification. MLID4 is a polysorbate surrogate and its conversion rate is an indicator for hydrolytic activity. Biological duplicates (n=2) were measured and are depicted as mean ± standard deviation. RFU = relative fluorescence units.

[0041]

[0038] Figure 14: Fluorescence micelle assay (FMA) with harvested cell culture fluid (HCCF) of lgG1 (A) and lgG4 (B) producing 9x hydrolase KO and parental CHO K1 cell pools after fed-batch cultivation. HCCF was diluted in fresh cultivation medium spiked with polysorbate (PS) and incubated for 0, 1 , 3, 7 and 14 days at room temperature, represented by 5 bars per sample from left to right in the diagram. Polysorbate degradation over time was measured in 4 technical replicates. Values were normalized to starting concentration on dO and are depicted as mean ± standard deviation.

[0042]

[0039] Figure 15: Ratio of high molecular weight (HMW) to monomer species in purified mAb fractions of lgG1 (A) and lgG4 (B) producing 9x hydrolase KO and parental CHO K1 cell pools after fed-batch cultivation. Purified mAb fraction was separated in HMWs and monomers using size exclusion chromatography and respective peak areas were calculated and are depicted in percentage of total area. Biological duplicates (n=2) were measured and are depicted as mean ± standard deviation.

[0043]

[0040] Figure 16: Ratio of low molecular weight (LMW) to monomer species in purified mAb fractions of lgG1 (A) and lgG4 (B) producing 9x hydrolase KO and parental CHO K1 cell pools after fed-batch cultivation. Purified mAb fraction was separated in LMWs and monomers using capillary electrophoresis and respective peak areas were calculated and are depicted in percentage of total area. Biological duplicates (n=2) were measured and are depicted as mean ± standard deviation.

[0044]

[0041] Figure 17: Charge variant analysis of purified mAb fractions of lgG1 (A) and lgG4 (B) producing 9x hydrolase KO and parental CHO K1 cell pools after fed-batch cultivation. Purified mAb fraction was separated in acidic (APG), neutral (Main) and basic charge variants (BPG) using ion exchange chromatography and respective peak areas were calculated and are depicted in percentage of total area. Biological duplicates (n=2) were measured and are depicted as mean ± standard deviation.

[0045]

[0042] Figure 18: Glycosylation pattern analysis of purified mAb fractions of lgG1 (A) and lgG4 (B) producing 9x hydrolase KO and parental CHO K1 cell pools after fed-batch cultivation. Purified mAb fraction was separated in its differently glycosylated species using capillary electrophoresis and respective peak areas were calculated and are depicted in percentage of total area. Biological duplicates (n=2) were measured and are depicted as mean ± standard deviation.

[0046]

[0043] Figure 19: (A) Genomic arrangement of carboxylesterase (Ces) cluster 1 in the parental CHO K1 host cell line. Filled arrows and / or boxes indicating genes on genome. Cas9:sgRNA cutting sites are indicated by arrows for the indicated sgRNAs (B) Schematic arrangement of Ces cluster 1 in a simplified depiction. Ces cluster 1 can be knocked out via sgRNA 1 / 2 and 3 delivering an approx. 1007 kilobase pair (kb) deletion or via sgRNA 1 / 2 and 4 delivering an approx. 1629kb deletion. Arrows indicating gene orientation on genome. (C) Gene expression in Ces cluster 1 of 189 CHO K1 datapoints representative for an industrially applied CHO cell line. Gene expression is depicted as transcript per million and represents mean ± standard deviation. (B, C) Gene variants are abbreviated with a serially numbered “V“; ? = not annotated genes; Cas9:sgRNA cutting sites are indicated by dashed lines.

[0047]

[0044] Figure 20: (A) Genomic arrangement of carboxylesterase (Ces) cluster 2 in the parental CHO K1 host cell line. (B) Schematic arrangement of Ces cluster 2 in a simplified depiction. Ces cluster 2 can be knocked out via sgRNA 1 and 2 delivering an approx. 506 kilobasepair (kb) deletion. Arrows indicating gene orientation on genome. (C) Gene expression in Ces cluster 2 of 189 CHO K1 datapoints representative for an industrially applied CHO cell line. Gene expression is depicted as transcript per million and represents mean ± standard deviation. Gene variants are abbreviated with a serially numbered “V“; ? = not annotated genes; Cas9:sgRNA cutting sites are indicated by arrows or dashed lines.

[0048]

[0045] Figure 21 : Fed-batch data from lgG1 (A, C and E) and lgG4 (B, D and F) producing 9x hydrolase + Bax / Bak1 (BB) KO cell pools. Parental host K1 clone and four 9x hydrolase + BB KO clones were stably transfected with lgG1 or lgG4 coding plasmids and resulting producer pools were compared in a fed-batch run. (A and B) Viable cell density (VCD) and viability [%] were measured from day 3 until harvest on day 14. (C and D) Volumetric productivity (Titer [mg / L]) and (E and F) specific productivity (qp [pg / cell*day]) were determined from day 7 to day 14. Cell pools were cultivated in biological duplicates (n=2) and data is shown as mean ± standard deviation.

[0049]

[0046] Figure 22: Fluorescence micelle assay (FMA) with harvested supernatant samples (day 14) from lgG1 (A) and lgG4 (B) producer fed-batch cultivations based on Bax / Bak1 (BB) multi-hydrolase KO and parental CHO K1 cell lines. Supernatant was diluted 1 :50 in fresh cultivation medium spiked with polysorbate (PS) and incubated for 0, 1 , 3, 7 and 14 days (d) at room temperature. PS degradation over time was measured in 4 technical replicates. Values were normalized to starting concentration on dO and are depicted as mean ± standard deviation. Hydrolytic activity was measured in supernatant (C) and protein A purified mAb fractions (D). Measurements were performed in biological duplicates and technical triplicates. Data is shown as mean ± standard deviation.

[0050]

[0047] Figure 23: IgG 1 and lgG4 product quality data of comparative fed-batch cultivations with parental and Bax / Bak1 -multi-hydrolase KO producer pools. IgG 1 and lgG4 of Bax / Bak1 (BB)-multi-hydrolase KO and parental CHO K1 producer pools were protein A purified and product quality was analyzed. (A / B) High molecular weight (HMW) ratios of lgG1 (A) and lgG4 (B) in comparison to monomer fraction analyzed via size exclusion chromatography; and (C / D) low molecular weight (LMW) ratios of lgG1 (C) and lgG4 (D) in comparison to monomer fraction analyzed via capillary gel electrophoresis. (E / F) Charge variant analysis of IgG 1 (E) and lgG4 (F) by ion exchange chromatography including acidic peak group (APG), main peak and basic peak group (BPG).

[0051]

[0048] Figure 24: Glycosylation pattern analysis of purified mAb fractions of lgG1 (A) and lgG4 (B) producing Bax / Bak1(BB)-multi-hydrolase KO and parental CHO K1 cell pools after fed-batch cultivation. Purified mAb fraction was separated in its differently glycosylated species using capillary gel electrophoresis and respective peak areas were calculated and are depicted in percentage of total area. * indicates the isomeric glycan variant. Biological duplicates (n=2) were measured and are depicted as mean ± standard deviation.

[0052] DETAILED DESCRIPTION

[0053]

[0049] The term “comprises” or “comprising” means “including, but not limited to”. The term is intended to be open-ended, to specify the presence of any stated features, elements, integers, steps, or components, but not to preclude the presence or addition of one or more other features, elements, integers, steps, components, or groups thereof. The term “comprising” thus includes the more restrictive terms “consisting of” and “essentially consisting of”. With regard to sequences, the terms “having a nucleotide / amino acid sequence of” and “comprising a nucleotide / amino acid sequence of” are used interchangeably and include the embodiment “consisting of the nucleic acid / amino acid sequence of’. Similarly, the term “encoding” or “encodes” is intended to be open-ended and allows the presence or addition of one or more other features, elements, or components. Furthermore, singular and plural forms are not used in a limiting way. As used herein, the singular forms “a”, “an” and “the” designate both the singular and the plural, unless expressly stated to designate the singular only.

[0054]

[0050] The term “protein” is used interchangeably with “amino acid sequence” or “polypeptide” and refers to polymers of amino acids of any length. These terms also include proteins that are post-translationally modified through reactions that include, but are not limited to, glycosylation, acetylation, phosphorylation, glycation, or protein processing. Modifications and changes, for example fusions to other proteins, amino acid sequence substitutions, deletions, or insertions, can be made in the structure of a polypeptide while the molecule maintains its biological functional activity. For example, certain amino acid sequence substitutions can be made in a polypeptide or its underlying nucleic acid coding sequence and a protein can be obtained with the same properties. A protein produced or manufactured in a mammalian cell or encoded by a mammalian expression vector, is also referred to as protein of interest. Preferably the protein of interest is a recombinant protein of interest, more preferably a therapeutic protein.

[0051] The term “recombinant protein” as used herein relates to a protein generated by recombinant techniques, such as molecular cloning and may also be referred to as recombinant protein of interest. It is coded for by a recombinant gene. As used herein, the recombinant protein is the protein of interest, e.g., in a sample to be purified or in a mammalian host cell (e.g., CHO cell) to be produced. Recombinant techniques bring together genetic material from multiple sources or create sequences that do not naturally exist. A recombinant protein is typically based on a sequence from a different cell or organism or a different species from the recipient host cell used for production of the protein in cell culture, e.g., a CHO cell, or is based on an artificial sequence, such as a fusion protein. In the context of the present invention the recombinant protein of interest is preferably a therapeutic protein, such as an antibody, an antibody fragment, an antibody derived molecule (e.g., scFv, bi- or multispecific antibodies) or a fusion protein (e.g., an Fc fusion protein). Thus, in one embodiment the recombinant protein of interest is selected from the group consisting of an antibody, an antibody fragment, an antibody derived molecule, glycoproteins and a fusion protein.

[0055]

[0052] The term “nucleic acid sequence” is used interchangeably with “polynucleotide” or “polynucleotide sequence” and refers to DNA or RNA of any length. In the context of an expression vector, particularly a plasmid and integration into the host cell’s genome, the person skilled in the art would understand that it refers to a DNA sequence or molecule. In the context of the present invention the nucleic acid sequence is typically a gene. In the context of a therapeutic protein to be encoded, the nucleic acid sequence is typically a recombinant gene.

[0056]

[0053] The term “encoding” or “coding for” as used herein refers to a nucleic acid molecule that has a sequence that provides the code for a specific amino acid sequence of a protein. Typically, a DNA sequence codes for a mRNA sequence and is therefore transcribed into a mRNA, while the mRNA codes for a protein and is therefore translated into an amino acid sequence of a protein. Encoding and coding for as used herein is an open language and may encompass further sequences. Thus, a plasmid or DNA encoding for (or coding for) a certain protein comprises a sequence coding for said protein, but may also code for a selection marker etc.

[0057]

[0054] The term “mammalian cell” as used herein refers to cells, particularly cell lines, derived from a mammal. In the present invention the host cell used is a Chinese Hamster Ovary cell (CHO cell) or cell clone or cell line derived thereof.

[0058]

[0055] The term “drug substance”, abbreviated as DS, as used herein refers to the formulated active pharmaceutical ingredient (API) with excipients. The API has the therapeutic effect in the body as opposed to the excipients, which assist with the delivery of the API. In the case of biologic therapeutics, the formulated API with excipients typically means the API in the final formulation buffer at a concentration of at least the highest concentration used in the final dosage form, also referred to as drug product.

[0059]

[0056] The term “drug product”, abbreviated as DP, as used herein refers to the final marketed dosage form of the drug substance for example a tablet or capsule or in the case of biologies typically the solution for injection in the appropriate containment, such as a vial or syringe. The drug product may also be in a lyophilized form. An antibody is preferably provided in an aqueous formulation in a glass vial or in a glass syringe.

[0060]

[0057] The term “polysorbate 20” as used herein refers to a non-ionic polysorbate-type surfactant, which is a laurate ester of sorbitol and its anhydrides, copolymerized with approximately 20 moles of ethylene oxide for each mole of sorbitol and sorbitol anhydrides (polyoxyethylene (20) sorbitan monolaurate; CAS number: 9005-64-5) as defined in ®2020 The United States Pharmacopeial Convention (Official May 1 , 2020). It is also known as Tween 20. Its stability and relative nontoxicity allow it to be used as a surfactant and emulsifier in a number of domestic, scientific analyses. Polysorbate 20 can be used as washing agent in immunoassays, Western blots and enzyme-linked immunosorbent assay (ELISA). It can further be used in pharmacological applications, such as pharmaceutical formulations, particularly for biologies, such as antibodies and Fc-fusion proteins. Particularly it helps to prevent non-specific antibody binding interactions.

[0061]

[0058] The term “polysorbate 80” as used herein refers to a non-ionic polysorbate-type surfactant, which is a mixture of partial esters of fatty acids mainly oleic acid, with sorbitol and its anhydrides ethoxylated with approximately 20 moles of ethylene oxide for each mole of sorbitol and sorbitol anhydrides (polyoxyethylene (20) sorbitan monooleate, CAS number: 9005-65-6) as defined in ®2015 The United States Pharmacopeial Convention (Stage 6 Harmonized, Official May 1 , 2016). It is also known as Tween 80 and has a similar use as polysorbate 20.

[0062]

[0059] The term “contaminating” or “contamination” as used herein refers to the presence of an undesired and / or unintentional substance, such as substances that are immunogenic and / or with PS degrading enzyme activity, e.g., enzymes with hydrolytic activities accompanying HCPs, particularly hydrolases in the context of the present invention, a hydrolytic activity is undesired due to its PS degrading potential that may be co-purified with the recombinant protein. This applies especially to finally formulated protein preparations which advantageously comprise such unwanted factors only to less than 1% (w / w), preferably less than 0.1% (w / w), more preferably less than 0.01% (w / w) in comparison to total protein content, i.e. , in the drug product.

[0060] The term “polysorbate degrading enzyme activity” or “polysorbate degrading activity” as used herein, refers to the activity of a substance, typically a hydrolase derived from the host cell, that catalyzes the hydrolysis of an ester bond in PS. This includes HPCs, specifically enzymes with hydrolytic activity, such as lipase activity and / or carboxylesterase activity. The term “lipase activity” as used herein refers to the activity of an enzyme that catalyzes the hydrolysis of an ester bond in a lipid, such as fatty acid esters. The term “carboxylesterase activity” as used herein refers to the activity of a substance, typically a protein (enzyme) that catalyzes the hydrolysis of an ester bond in a carboxylic ester. A lipase or a carboxylesterase is a hydrolase enzyme that splits esters into an acid and an alcohol in a chemical reaction with water, also referred to as hydrolysis. Many lipases and carboxylesterases belong to the class of carboxylic ester hydrolases (EC 3.1.1). While carboxylesterases form a separate class of carboxylesterase (EC 3.1.1.1), lipases may be, without being limited thereto, a triacylglycerol lipase (EC 3.1.1.3), a phospholipase A2 (EC

[0063] 3.1.1.4), a lysophospholipase (EC 3.1.1.5), an (EC 3.1.1.23), galactolipase (EC 3.1.1.26), phospholipase A1 (EC 3.1.1.32), lipoprotein lipase (EC 3.1.1.34) or hormone-sensitive lipase (EC 3.1.1.79); a phosphoric diester hydrolase (EC 3.1.4) such as phospholipase D (EC

[0064] 3.1.4.4), a phosphoinositide phospholipase C (EC 3.1.4.11), glycosylphosphatidylinositol phospholipase D (EC 3.1.4.50) or N-acetylphosphatidylethanolamine-hydrolyzing phospholipase D (EC 3.1.4.54); or a glycosphingolipid deacylase (EC 3.5.1.69). The term “hydrolase activity” as used herein is the more general term, including lipases and carboxylesterases, but also refers to hydrolysis of compounds other than lipids, such as the sialic acid acetylesterase (SIAE) that catalyze removal of O-acetyl ester groups from sialic acids by hydrolysis and thioester hydrolase (EC 3.1.2), such as palmitoyl protein thioesterase 1 (PPT1). Of particular interest in the context of the present invention are hydrolases acting on ester bonds. The term “hydrolase acting on ester bonds” as used herein relates to enzymes (particularly lipases or carboxylesterases) having hydrolase activity that catalyze the hydrolysis of an ester bond, wherein “acting on ester bonds” means cleaving ester bonds.

[0065]

[0061] The term “hydrolase gene” as used herein refers to a gene encoding a hydrolase acting on ester bonds, particularly a hydrolase acting on the ester bond in PS (e.g., polysorbate 20 and / or polysorbate 80). PS degrading hydrolases were characterized herein to comprise a serine (Ser) - histidine (His) - aspartate / glutamate (Asp / Glu) motif (also referred to as catalytical triad), which distinguishes them from non-PS degrading enzymes, and preferably belong to the hydrolase class with the EC number 3.1. For example, acid ceramidase comprising the same catalytic triad and belonging to the hydrolase class with the EC number 3.5 does not exhibit PS degrading activity. Thus, the catalytic triad is a prerequisite for PS degradation, but other protein properties also seem to be important.

[0062] The term “therapeutic protein” as used herein refers to proteins that can be used in medical treatment of humans and / or animals. Therapeutic proteins are typically recombinant proteins and include, but are not limited to antibodies, growth factors, blood coagulation factors, vaccines, interferons, hormones, and fusion proteins. In the context of the present invention the therapeutic protein is an antibody, including an antibody-derived molecule.

[0066]

[0063] The term “produced” or “producing” as used herein relates to the production of a recombinant protein of interest, specifically an antibody, in a CHO cell in cell culture. The person skilled in the art knows how to produce recombinant proteins and particularly antibodies in cells using fermentation. The production of recombinant proteins comprises cultivating a mammalian cell, such as a CHO cell, expressing the recombinant protein of interest in cell culture. Cultivating the cell expressing the recombinant protein in cell culture comprises maintaining the cell in a suitable medium and under conditions that allow growth and / or protein production / expression. The recombinant protein of interest may be produced by fed-batch or continuous cell culture. Thus, the CHO cell may be cultivated in a fed-batch or continuous cell culture or a combination thereof, preferably in a fed-batch cell culture.

[0067]

[0064] The term “expressing a recombinant protein of interest” or “expressing an antibody” as used herein refers to a CHO cell comprising a DNA sequence coding for the recombinant protein of interest, such as an antibody, which is transcribed and translated into the protein sequence including post-translational modifications, i.e., resulting in the production of the recombinant protein of interest, such as an antibody in cell culture.

[0068]

[0065] The term “about” as used herein refers to a variation of 10 % of the value specified, for example, about 50 % comprises a variation from 45 to 55 %.

[0069]

[0066] The term “two or more” and “at least two” or equivalents are used synonymously herein and refer to two or more than two. The same applies to equivalents for other numbers, such as three or more or four or more.

[0070]

[0067] The term “yield” as used herein refers to the amount of the antibody secreted into the harvested cell culture fluid (HCCF) and / or following purification relative to the amount of the antibody before purification, such as in the starting material.

[0071]

[0068] The term "expression" as used herein refers to transcription and / or translation of a heterologous polynucleotide sequence within a host cell. The level of expression of a polynucleotide sequence, such as a polynucleotide sequence encoding a hydrolase acting on ester bonds or a protein of interest in a host cell may be determined based on either the amount of corresponding mRNA that is present in the cell, or the amount of the desired polypeptide / protein of interest encoded by the selected sequence as in the present examples. For example, mRNA transcribed from a selected sequence can be quantified by Northern blot hybridization, ribonuclease RNA protection, in situ hybridization to cellular RNA or by polymerase chain reaction (PCR). Proteins encoded by a selected sequence can be quantitated by various methods, e.g. by ELISA, by Western blotting, by radioimmunoassays, by immunoprecipitation, by assaying for the biological activity of the protein, by immunostaining of the protein followed by flow cytometry analysis or by homogeneous time- resolved fluorescence (HTRF) assays. The level of expression of a non-coding RNA, such as a miRNA can be quantified by PCR, such as qPCR.

[0072]

[0069] The term “harvested cell culture fluid” as used herein refers to the cell culture fluid (CCF) from which the cells have been removed during harvest, typically by centrifugation or filtration. Thus, the harvested cell culture fluid refers to the cell culture supernatant following harvest, i.e., cell removal, particularly for secreted proteins.

[0073] Multiple-hydrolase knockout CHO cells

[0074]

[0070] The present invention identifies the most critical hydrolases for PS degradation in compositions comprising recombinant proteins produced in CHO cells and provides and generates CHO cells with these hydrolases being knocked out. In addition to functional gene deletion for most hydrolases, for Ces1 , Ceslf and Ces2c genes, which are localized in gene clusters, the entire gene clusters were deleted. This large excision of the entire gene clusters removed all potential Ces variants, thereby avoiding putative functional reactivation of related Ces species. To the best of our knowledge this excision of > 1 megabases is the largest excision in CHO cells to date. Surprisingly the generated cells showed cell culture parameters (VCD, viability) and productivity (titer, specific productivity) suitable for pharmaceutical (large-scale) production without negatively affecting product quality attributes. Hydrolase activity was shown to gradually decrease with each knockout. Particularly knockout of the two Ces clusters reduced PS degradation in HCCF, which was even further reduced by the following gene knockouts. This provides for the first time a CHO cell line allowing recombinant protein production with neglectable hydrolase activity due to an up to 9x hydrolase KO and with only slightly reduced, but acceptable viability and VCD. Viability and VCD can further be rescued by additional Bax / Bak1 gene knockouts. This cell line is therefore well suitable for recombinant protein production, particularly therapeutic protein production, such as antigens or Fc protein, particularly for more complex proteins, such as antibody-derived molecules, bispecific antibodies and also single-chain formats, for which optimization of purification trains is even more difficult due to their diverse physiochemical properties.

[0075]

[0071] The present invention provides a CHO cell comprising, (i) a lipoprotein lipase (LPL) gene knockout (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene cluster comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) of the CHO genome and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). A gene cluster is a group of two or more genes belonging to the same gene family, e.g., that encode similar polypeptides or proteins, which share a similar function and / or sequence and are located within a few thousand base pairs of each other (e.g., within 500 to 1000 kb). Typically, a gene cluster has 5 or more, 10 or more, 15 or more or even 20 or more genes belonging to the same gene family, i.e., that encode similar polypeptides or proteins. The term “two or more genome excisions deleting at least two gene cluster comprising at least one carboxylesterase gene” as used herein means that the at least two gene clusters each comprising at least one carboxylesterase gene are deleted or excised, however, in addition other genome excisions may be present, such as resulting in a total of 3, 4, 5 or more genome excisions, e.g., 2 to 10 or 2 to 5 genome excisions. The term “at least two gene clusters comprising at least one carboxylesterase gene” or “a gene cluster comprising at least one carboxylesterase gene” specifies that each of the gene cluster comprises at least one carboxylesterase gene (or 5 or more, 10 or more, 15 or more or 20 or more). This defines that the gene cluster is a carboxylesterase gene cluster, such as the CES1 cluster or the CES2 cluster. Thus, the CHO cell comprises two or more gene excisions comprising a first genome excision of a first gene cluster and a second genome excision of a second gene cluster, wherein the first gene cluster and the second gene cluster each comprises at least one carboxylesterase gene, which means a first genome excision of a first carboxylesterase gene cluster and a second genome excision of a second carboxylesterase gene cluster.

[0076]

[0072] Preferably the hydrolase genes that are knocked-out and the at least one carboxylesterase gene comprised in each of the two or more gene clusters that are excised code for hydrolases and carboxylesterases that have been identified in a drug substance or drug product and that have been confirmed to degrade polysorbates (PS), particularly polysorbate 20 (PS20) and polysorbate 80 (PS80). Carboxylesterases form a separate class of hydrolases (EC Number 3.1.1.1) and include without being limited thereto CES1 , CES1 F and CES2C, which have been identified in drug substances or drug products of specific antibodies produced in CHO cells and that have been confirmed to degrade polysorbates.

[0077]

[0073] While a gene knockout comprises gene excision and gene inactivation, a genome excision clearly defines a deletion of a consecutive genomic DNA strand. Both, knockout (gene deletion / excision and gene inactivation) and genome excision requires targeted genome editing. Means for targeted genome editing are known in the art and include without being limited thereto zinc finger nucleases (ZNFs), transcription activator- 1 ike effector nucleases (TALEN) and clustered regulated interspaced short palindromic repeats (CRISPR), particularly CRISPR / Cas9. Gene knockouts can be mediated by non-homologous end joining (NHEJ) or homology directed repair (HDR). Both mechanisms rely on cellular repair mechanisms that either accurately delete a gene or part of a gene requiring a template with homologous sequences (HDR) or rejoin broken DNA ends without a template with homologous sequences by inducing insertion and / or deletions (InDei) in the target gene (NHEJ). The latter is ideal for gene knockouts where disrupting gene function is the goal.

[0078]

[0074] According to certain embodiments the LPL knockout, the at least three hydrolase knockouts and the two or more genome excisions may be biallelic. The person skilled in the art will understand that the knockouts and / or genome excisions do not need to be strictly homozygous, since for example the InDei may differ between the alleles or the genome excision may be slightly larger on one allele then in the other allele.

[0079]

[0075] In contrast to gene inactivation (i.e., disruption of gene function), such as by InDei mutations, gene excision or genome excision minimizes the genome. Particularly genome excision of a gene cluster removing a few thousands of base pairs minimizes the genome. According to the invention each genome excision removes at least 500 kilobase pairs (kb) of the CHO genome, such as 500 kb to 10.000 kb or 500 kb to 5000 kb or 500 kb to 2000 kb, preferably 1000 to 10.000 kb or 1000 to 5000 kb or 1000 kb to 2000 kb. In certain embodiments at least one of the two or more genome excisions removes 1000 kb or more, preferably 1500 kb or more of the CHO genome, such as 1000 kb to 10.000 kb, preferably 1500 kb to 10.000 kb or more of the CHO genome or 1000 kb to 5.000 kb or 1000 kb to 2000 kb, preferably 1500 kb to 5.000 kb or 1500 kb to 2000 kb of the CHO genome.

[0080]

[0076] According to the invention the CHO cell comprises (ii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kb of the CHO genome and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). This means the CHO cell comprises (ii) two or more genome excisions comprising genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C), wherein each genome excision removes at least 500 kb of the CHO genome. In certain embodiments, the two or more genome excisions comprise or consist of genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). In a more specific embodiment, the two or more genome excisions are genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). Thus, the first gene cluster comprises at least one carboxylesterase gene encoding CES1 comprising the sequence of SEQ ID NO: 2 and / or CES1 F comprising the sequence of SEQ ID NO: 4 or a sequence comprising at least 90%, preferably at least 95%, sequence identity with SEQ ID NOs: 2 or 4 and the second gene cluster comprises at least one carboxylesterase gene encoding CES2C comprising the sequence of SEQ ID NO: 1 or a sequence comprising at least 90%, preferably at least 95%, sequence identity with SEQ ID NO: 1. In preferred embodiments the two or more genome excisions comprise genome excisions (a) of the CES1 cluster removing at least 1000 kb and (b) of the CES2 cluster removing at least 500 kb. In certain embodiments the two or more genome excisions comprise genome excisions (a) of the CES1 cluster removing 1000 kb to 10.000 kb, preferably 1500 kb to 10.000 kb or 1000 kb to 5.000 kb, more preferably 1500 kb to 5.000 kb or 1000 kb to 2000 kb of the CHO genome, and (b) of the CES2 cluster removing 500 kb to 10.000 kb, preferably 500 kb to 5000 kb or 1000 to 10.000 kb, more preferably 1000 to 5000 kb. The CES1 cluster is comprised in the nucleic acid sequence of SEQ ID NO: 41 (1004 kb) and in the longer nucleic acid sequence of SEQ ID NO: 31 (1629 kb). Thus, in a specific embodiment, the genome excision (a) of the CES1 cluster removing at least 1000 kb comprises a nucleic acid sequence corresponding to SEQ ID NOs: 41 or 31. The CES2 cluster is comprised in the nucleic acid sequence of SEQ ID NO: 42 (504 kb) and in the longer nucleic acid sequence of SEQ ID NO: 32 (506 kb). Thus, in a specific embodiment, the genome excision (a) of the CES2 cluster removing at least 500 kb comprises a nucleic acid sequence corresponding to SEQ ID NOs: 42 or 32, such as 500 kb to 10.000 kb or 500 kb to 5000 kb or 500 kb to 2000 kb, preferably 1000 to 10.000 kb or 1000 to 5000 kb or 1000 kb to 2000 kb comprising the nucleic acid sequence corresponding to SEQ ID NOs: 42 or 32. The person skilled in the art will understand that the term “removing at least 1000 kb” or the like refers to removal of at least 1000 kb of the CHO genome, i.e. , to a genomic DNA strand that is at least 1000 kb long. The term ’’comprising a nucleic acid sequence corresponding to SEQ ID NO: X” as used herein defines the sequence even if slightly different in a different CHO cell line. Typically, CHO genome sequences have more than about 98% sequence identity to each other, preferably more than about 99% sequence identity to each other. The terms “CES1 cluster” and “Ces cluster 1” are used synonymously herein and can be exchanged. Similarly, the terms “CES2 cluster” and “Ces cluster 2” are used synonymously herein and can be exchanged.

[0077] In specific embodiments, the two or more genome excisions deleting at least two gene clusters comprising a carboxylesterase gene additionally comprises upstream and / or downstream flanking genomic regions, wherein the upstream and / or downstream flanking genomic regions are preferably 200 kb, 300 kb or 600 kb or more, such as 200-1000 kb, 300- 1000 kb or 600-1000 kb. The person skilled in the art will understand that each individually or both of the two or more genome excisions may further excise upstream and / or downstream flanking genomic regions. The CHO cell according to the invention may further comprise at least one additional genome excision of a gene cluster and / or of at least 800 kb of the CHO genome. Other suitable gene clusters are, without being limited thereto, protocadherin clusters, which are organized in three closely linked gene clusters designated the protocadherin alpha (Pcdha) cluster, the protocadherin beta (Pcdhb) cluster and the protocadherin gamma (Pcdhg) cluster.

[0081]

[0078] The at least three further hydrolases gene knockouts are knockouts of genes coding for hydrolases that have been identified in a drug substance or purified antibody composition and have polysorbate degrading activity (PS20 and / or PS80). Typically, such hydrolases are of the hydrolase class with the EC number 3.1 and comprise a catalytical triad consisting of a serine (Ser)-histidine (His)-aspartate / glutamate (Asp / Glu) motif. The term “at least three further hydrolase gene knockouts” includes without being limited thereto, e.g., 3 to 10, or 4 to 10, or 5 to 10, preferably 4 to 8, more preferably 5 to 8 further hydrolase gene knockouts, but may also include as many further hydrolases gene knockouts as may be knocked-out maintaining (or without substantially affecting) cell growth, viable cell density and titer. In certain embodiments of the CHO cell of the invention the (ii) at least three further hydrolase gene knockouts comprise at least three knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH 1 ) , lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2). In alternative or additional embodiments the (ii) at least three further hydrolase gene knockouts comprise at least three knockouts of genes coding for hydrolases of the hydrolase class with the EC number 3.1 and comprising a catalytical triad consisting of a serine (Ser)-histidine (His)-aspartate / glutamate (Asp / Glu) motif. In more specific embodiments the (ii) at least three further hydrolase gene knockouts comprise knockouts of genes coding for hydrolases (a) selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein- thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2), and / or (b) of the hydrolase class with the EC number 3.1 and comprising a catalytical triad consisting of a serine (Ser)-histidine (His)- aspartate / glutamate (Asp / Glu) motif. Preferably, the (ii) at least three further hydrolase gene knockouts comprise knockouts of hydrolase genes selected from the group consisting of lipase A (LIPA), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2).

[0082]

[0079] In certain embodiments, the CHO cell of the invention comprises (i) a lipoprotein lipase (LPL) gene knockout, (ii) at least five further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters (each) comprising at least one carboxylesterase gene, specifically comprising genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). In certain embodiments, the (ii) at least five further hydrolase gene knockouts comprise at least three knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2). In more specific embodiments, the (ii) at least five further hydrolase gene knockouts comprise knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH 1 ), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2). In certain embodiments, the CHO cell according to the invention comprising a recombinant gene encoding a therapeutic protein has an increased specific productivity (qp [pg / cell / day]). The term “increased” as used in this context means an increased specific productivity compared to a control CHO cell (reference cell) not comprising (i) a lipoprotein lipase (LPL) gene knockout, the (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb), and / or comprising (i) a lipoprotein lipase (LPL) gene knockout, but without comprising the (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb). In certain embodiments the specific productivity (qp [pg / cell / day]) is increased by at least 20%, at least 50%, at least 80% or even higher.

[0083]

[0080] The recited hydrolases including the gene names, synonyms and the respective SEQ ID NOs are summarized in Table 1. Table 1 : Summary of knocked out polysorbate degrading enzymes with information on hydrolase name, abbreviations for gene and protein, synonyms and Uniprot accession

[0084]

[0081] The CHO cell according to the invention may further comprise additional gene knockouts, such as a gene knockout of the gene coding for alpha-(1 ,6)-fucosyltransferase (FLIT8), a gene knockout of the gene coding for glutamine synthetase (GS) and / or a gene knockout of the genes coding for BAX and BAK. Particularly gene knockout of the genes coding for BAX and BAK (e.g., amino acid sequences of SEQ ID NO: 26 and 27) may be advantageous to increase cell viability, such as in case of a critical reduction in cell viability in a production process, e.g. a reduced cell viability at the end of the production process (below 60-80% viable cells), and / or compared to a control CHO cell (or the CHO cell comprising (i) a LPL gene knockout) not comprising the (ii) the at least three further hydrolase gene knockouts and (iii) the two or more genome excisions according to the invention. Thus, in certain embodiments the CHO cell further comprises a gene knockout of the genes coding for BAX and BAK.

[0085]

[0082] Since the GS gene is a selectable marker gene frequently used for CHO cell line development, the CHO cell according to the invention may be a CHO GS knockout cell. Thus, in preferred embodiments the CHO cell further comprises at least a gene knockout of the gene coding for GS, more preferably additionally a gene knockout for the gene coding for FLIT8 and / or BAX and BAK. In a more preferred embodiment, the CHO cell further comprises a gene knockout of the genes coding for GS, BAX and BAK. These cells may optionally further comprise a gene knockout for the gene coding for FLIT8.

[0086]

[0083] The term “gene knockout of the gene coding for GS” and “gene knockout of the genes coding for BAX and BAK” is used synonymously herein with “Glul gene knockout” and “Bax and Bak1 gene knockout” or “GS knockout” and “BAX and BAK knockout”, respectively.

[0087]

[0084] The recited additional gene knockout, including the gene names, synonyms and the respective SEQ ID NOs are summarized in Table 2.

[0088] Table 2: Non-limiting examples of other knocked out targets in CHO host cells with information on target name, abbreviations for gene and protein, synonyms and Uniprot

[0089]

[0085] The CHO cell according to the present invention may further comprise a recombinant gene encoding a therapeutic protein. Typically, the CHO cell expressing a therapeutic protein also comprise a selection marker gene.

[0090]

[0086] A “selectable marker gene” or “selection marker gene” is a gene which encodes a selectable marker and allows the specific selection of cells which contain this gene, typically by the addition of a corresponding “selecting agent” to the cultivation medium. As an illustration, an antibiotic resistance gene may be used as a positive selectable marker. Only cells which have been transformed with this gene are able to grow in the presence of the corresponding antibiotic and are thus selected. Untransformed cells, on the other hand, are unable to grow or survive under these selection conditions. There are positive, negative and bifunctional selectable markers. Positive selectable markers permit the selection and hence enrichment of transformed cells by conferring resistance to the selecting agent or by compensating for a metabolic or catabolic defect in the host cell. By contrast, cells which have received the gene for the selectable marker can be selectively eliminated by negative selectable markers. An example of this is the thymidine kinase gene of the Herpes Simplex virus, the expression of which in cells with the simultaneous addition of acyclovir or ganciclovir leads to the elimination thereof. The selectable marker genes useful in this invention also include the amplifiable selectable markers. The literature describes a large number of selectable marker genes including bifunctional (positive / negative) markers (see for example WO 92 / 08796 and WO 94 / 28143). Examples of selectable markers which are useful in the present invention include, but are not limited to the genes of aminoglycoside phosphotransferase (APH), hygromycine phosphotransferase (HYG), dihydrofolate reductase (DHFR), thymidine kinase (TK), glutamine synthetase, asparagine synthetase and genes which confer resistance to neomycin (G418 / Geneticin), puromycin, histidinol D, bleomycin, phleomycin, blasticidin and zeocin. Glutamine synthetase is particularly preferred. Also included are genetically modified mutants and variants, fragments, functional equivalents, derivatives, homologues and fusions with other proteins or peptides, provided that the selectable marker retains its selective qualities. Such derivatives display considerable homology in the amino acid sequence in the regions or domains, which are deemed to be selective.

[0091]

[0087] In another aspect, the invention provides a method of producing a recombinant protein, comprising, (a) introducing a nucleic acid comprising a gene coding for a therapeutic protein and a gene coding for a selection marker into the CHO cell according to the invention; (b) culturing the cell of step (a) under conditions to produce the therapeutic protein; (c) harvesting the therapeutic protein; and (d) purifying and optionally formulating the therapeutic protein. In certain embodiments the method provides a comparable or increased specific productivity (qp), and preferably without affecting product quality (e.g., glycosylation, charge species, high molecular weight species, low molecular weight species etc.) compared to the same method using a control CHO cell (reference cell) not comprising (i) a lipoprotein lipase (LPL) gene knockout (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb), and / or comprising (i) a lipoprotein lipase (LPL) gene knockout, but without comprising (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb). In certain embodiments the method provides a comparable or increased titer compared to the same method using a control CHO cell (reference cell) not comprising (i) a lipoprotein lipase (LPL) gene knockout (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb), and / or comprising (i) a lipoprotein lipase (LPL) gene knockout, but without comprising (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb). The person skilled in the art will understand that the control CHO cell particularly does not comprise two or more genome excisions comprising genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C).

[0092]

[0088] The introducing in step (a) may be transfection or transduction, preferably transfection, wherein the nucleic acid is stably integrated into the CHO genome. Moreover, the gene coding for the therapeutic protein and the gene coding for the selection marker are typically part of the same DNA molecule, such as a plasmid. In certain embodiments, the selection marker is GS.

[0093]

[0089] In yet another aspect, the invention provides a method of reducing polysorbate degrading activity in a CHO cell, comprising (a) providing a CHO cell; (b) knocking out (i) lipoprotein lipase (LPL) gene and (ii) at least three further hydrolase gene knockouts in the CHO cell; and (c) excising genomic regions comprising (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) in the CHO cell, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). This method produces a CHO production cell line suitable for the production of a recombinant protein.

[0094]

[0090] In yet another aspect, the invention relates to a method of reducing polysorbate degrading activity in a recombinant protein produced in a CHO cell, wherein the method comprises (a) providing a CHO cell; (b) knocking out (i) lipoprotein lipase (LPL) gene and (ii) at least three further hydrolase gene knockouts in the CHO cell; (c) excising genomic regions comprising (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) in the CHO cell, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C); (d) transfecting the CHO cell comprising the knockout of step (b) and the two or more genome excisions of step (c) with a gene coding for a recombinant protein, preferably a therapeutic protein and a gene coding for a selection marker; (e) harvesting the therapeutic protein; and (f) purifying and optionally formulating the therapeutic protein. The transfection in step (d) is preferably a stable transfection and more preferably the gene coding for the recombinant protein and the gene coding for the selection marker are part of the same DNA molecule. Typically, the DNA molecule is a plasmid. In certain embodiments the selection marker is GS.

[0095]

[0091] In certain embodiments, the methods according to the invention provide a reduced polysorbate degrading activity in a drug substance or drug product compared to the same method using a control CHO cell (reference cell) not comprising (i) a lipoprotein lipase (LPL) gene knockout the (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb), and / or comprising (i) a lipoprotein lipase (LPL) gene knockout, but without comprising the (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) (wherein the CHO control cells are otherwise the same cells). The person skilled in the art will understand that the control cell particularly does not comprise two or more genome excisions comprising genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C). Wherein reduced means reduced by at least 20%, preferably at least 50%, preferably at least 75%, preferably at least 80%, preferably at least 90%, at least 95% or even by at least 95% compared to a method using a control CHO cell as specified herein. Wherein the polysorbate degrading activity may be measured using the FMA assay or the MUD4 / MUN4 assay as described herein, preferably using the FMA assay. The person skilled in the art will understand that the polysorbate degrading activity is measured using the same assay and the same conditions as for the control CHO cells.

[0096]

[0092] In certain embodiments, the CHO cell according to the invention comprising a recombinant gene encoding a therapeutic protein has an increased specific productivity (qp [pg / cell / day]). The term “increased” as used in this context means an increased specific productivity compared to a control CHO cell (reference cell) not comprising (i) a lipoprotein lipase (LPL) gene knockout (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb), and / or comprising (i) a lipoprotein lipase (LPL) gene knockout, but without comprising (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb).

[0097]

[0093] In yet another aspect the method relates to a pharmaceutical composition comprising a recombinant protein produced according to the methods of the invention, optionally further comprising pharmaceutically acceptable excipients.

[0098]

[0094] Reduced hydrolase activity may be determined, without being limited thereto, by means of a fluorescent assay as for example described in WO 2022 / 049294 A1 for detecting lipase activity using 4-methylumbelliferyl decanoate (4-MUD or MLID4) or 4- methylumbelliferyl nonanoate (4-MUN or MLIN4) as substrate and / or in WO 2023 / 170139 A1 for detecting hydrolases such as carboxylesterases using 1-octanoyloxy-pyrene-3,6,8- trisulfonic acid (OPTS) as substrate. PS concentrations may be determined, e.g., using the fluorescence micelle assay (FMA) as described herein. Thus, for determining reduced PS degrading activity samples, such as HOOF samples (diluted), protein A purified antibody samples or drug substance samples, may be spiked with PS20 (e.g., cPS20 = 0.4 mg / ml) or PS80 (e.g., cPS80 = 0.2 mg / ml) following storage at room temperature (about 22°C) at day 0, 1 , 3, 7 and 14, and frozen at -70°C prior to analysis using the FMA. For the FMA 10 pl sample are mixed with 240 pl FMA reagent (150 mM NaCI, 50 mM Tris, 0.2% Acetonitrile, 5 pM NPN, 0.0015% Brij-35, pH 8) and subsequently incubated for 1 min at 35°C. Fluorescent signal, positively correlating with PS concentration, may be measured at 420 nm following excitation of the dye with a 350 nm laser using a microplate reader (such as from Infinite® 200 PRO Tecan, Mannedorf, Switzerland). PS concentrations are quantified using standard curves for PS20 (such as ranging from 0.1 to 0.6 mg / ml) and / or for PS80 (such as ranging from 0.05 to 0.3 mg / ml). Reduction compared to control may be determined by comparing PS20 or PS80 in mg / ml at various days or by comparing the degradation rate (mg / ml / day).

[0099]

[0095] The term “4-MUD / 4-MUN assay” and “as measured by 4-MUD / 4-MUN assay” as used herein refers to measuring lipase activity using 4-methyumbelliferyl decanoate (4-MUD) or 4-methylumbelliferyl nonanoate (4-MUN) as substrate as described herein, particularly measuring lipase activity in a reaction mixture comprising 3-30 pM 4-MUD or 4-MUN in AMT assay buffer (75 mM acetate, 75 mM MES, 150 mM Tris, 150 mM NaCI, 10 mM CHAPS, at the indicated pH, preferably at pH 5.5) and preferably detected using a microplate reader (Aem = 450 nm, Aex = 330 nm, top read mode) in a black 96 well plate. The composition preferably comprises the recombinant protein at 50-200 mg / ml, such as 150 mg / ml, and may be measured at sample volume of 72 pl in a total volume of the reaction mixture of 300 pl, e.g. at 12-50 mg / ml, such as 36 mg / ml of a recombinant protein (e.g., an antibody).

[0096] The CHO cell according to the present invention may further express a recombinant protein of interest, preferably a therapeutic protein, more preferably a therapeutic antibody. The recombinant protein of interest may be stably or transiently expressed, preferably the recombinant protein of interest is stably expressed.

[0100]

[0097] The recombinant protein of interest may be any protein and is typically a therapeutic protein. The term “therapeutic protein” as used herein refers to proteins that can be used in medical treatment of humans or animals. These include, but are not limited to cytokines, growth factors, hormones, blood coagulation factors, vaccines, interferons, fusion proteins, antibodies, antibody-derived molecules, and an antibody mimetic. In certain embodiments, the therapeutic protein is selected from the group consisting of a cytokine, a hormone, a fusion protein, an antibody (including an antibody-derived molecule or antibody fragments), and an antibody mimetic.

[0101]

[0098] A preferred recombinant protein of interest is an antibody, including fragments and derivatives thereof. Typically, an antibody is monospecific, but an antibody may also be multispecific (symmetric or asymmetric). Thus, the present invention may be used for the production of monospecific antibodies, multispecific antibodies, or fragments thereof, preferably of antibodies (monospecific), bispecific antibodies, trispecific antibodies or fragments or combinations thereof, preferably antigen-binding fragments thereof. Exemplary antibodies within the scope of the present invention include but are not limited to anti-CD2, anti-CD3, anti-CD20, anti-CD22, anti-CD30, anti-CD33, anti-CD37, anti-CD40, anti-CD44, anti-CD44v6, anti-CD49d, anti-CD52, anti-EGFR1 (HER1), anti-EGFR2 (HER2), anti-GD3, anti-IGF, anti-VEGF, anti-TNFalpha, anti-IL2, anti-IL-5R or anti-lgE antibodies, and are preferably selected from the group consisting of anti-CD20, anti-CD33, anti-CD37, anti- CD40, anti-CD44, anti-CD52, anti-HER2 / neu (erbB2), anti-EGFR, anti-IGF, anti-VEGF, anti- TNFalpha, anti-l L2 and anti-lgE antibodies.

[0102]

[0099] The term “antibody”, “antibodies”, or“immunoglobulin(s)” is used herein in the broadest sense and encompasses various antibody structures, including but not limited to monoclonal antibodies, polyclonal antibodies, monospecific antibodies, multispecific antibodies (e.g., bispecific antibodies), antibody-derived molecules and antibody fragments so long as they exhibit the desired antigen-binding activity. There are various classes of immunoglobulins: IgA, IgD, IgE, IgG, IgM, IgY, IgW. Preferably the antibody is an IgG antibody, more preferably an IgG 1 or an lgG4 antibody.

[0103]

[0100] Antibodies can be of any species and include chimeric and humanized antibodies. “Chimeric” antibodies are molecules in which antibody domains or regions are derived from different species. For example, the variable region of heavy and light chain can be derived from rat or mouse antibody and the constant regions from a human antibody. In “humanized” antibodies only minimal sequences are derived from a non-human species. Often only the complementarity-determining region (CDR) amino acid residues of a human antibody are replaced with the CDR amino acid residues of a non-human species such as mouse, rat, rabbit or llama. Sometimes a few key framework amino acid residues with impact on antigen binding specificity and affinity are also replaced by non-human amino acid residues.

[0104]

[0101] Typically, antibodies are tetrameric polypeptides composed of two pairs of a heterodimer each formed by a heavy and a light chain. Stabilization of both the heterodimers as well as the tetrameric polypeptide structure occurs via interchain disulfide bridges. Each chain is composed of structural domains called “immunoglobulin domains” or “immunoglobulin regions” whereby the terms “domain” or “region” are used interchangeably. Each domain contains about 70 - 110 amino acids and forms a compact three-dimensional structure. Both heavy and light chain contain at their N-terminal end a “variable domain” or “variable region” with less conserved sequences which is responsible for antigen recognition and binding. The variable region of the light chain is also referred to as “VL” and the variable region of the heavy chain as “VH”.

[0105]

[0102] An “antibody fragment” or “antigen-binding fragments” refers to a molecule other than an intact antibody that comprises a portion of an intact antibody that binds the antigen to which the intact antibody binds. Examples of antibody fragments include but are not limited to Fv, Fab, Fab’, Fab’-SH, F(ab’) 2; diabodies; linear antibodies; single-chain variable fragment antibody molecules (e.g. scFv); and multispecific antibodies formed from antibody fragments. Fab fragments consist of the variable regions of both chains, which are held together by the adjacent constant region. These may be formed by protease digestion, e.g., with papain, from conventional antibodies, but similarly Fab fragments may also be produced by genetic engineering. Further antibody fragments include F(ab‘)2 fragments, which may be prepared by proteolytic cleavage with pepsin.

[0106]

[0103] Using genetic engineering methods, it is possible to produce shortened antibody fragments which consist only of the variable regions of the heavy (VH) and of the light chain (VL). These are referred to as Fv fragments (Fragment variable = fragment of the variable part). Since these Fv-fragments lack the covalent bonding of the two chains by the cysteines of the constant chains, the Fv fragments are often stabilized. It is advantageous to link the variable regions of the heavy and of the light chain by a short peptide fragment, e.g. of 10 to 30 amino acids, preferably 15 amino acids. In this way a single peptide strand is obtained consisting of VH and VL, linked by a peptide linker. An antibody protein of this kind is known as a scFv. Examples of scFv-antibody proteins are known to the person skilled in the art. Thus, antibody fragments and antigen-binding fragments further include Fv-fragments and particularly scFv.

[0104] In recent years, various strategies have been developed for preparing scFv as a multimeric derivative. This is intended to lead, in particular, to recombinant antibodies with improved pharmacokinetic and biodistribution properties as well as with increased binding avidity. In order to achieve multimerisation of the scFv, scFv were prepared as fusion proteins with multimerisation domains. The multimerisation domains may be, e.g. the CH3 (CH3 = 3rdconstant heavy chain domain) region of an IgG or coiled coil structure (helix structures) such as Leucine-zipper domains. However, there are also strategies in which the interaction between the VH / VL regions of the scFv is used for the multimerisation (e.g. dia-, tri- and pentabodies). By diabody the skilled person means a bivalent homodimeric scFv derivative. The shortening of the linker in a scFv molecule to 5 - 10 amino acids leads to the formation of homodimers in which an inter-chain VH / VL-superimposition takes place. Diabodies may additionally be stabilized by the incorporation of disulfide bridges. Examples of diabody- antibody proteins are known from the prior art.

[0107]

[0105] By minibody the skilled person means a bivalent, homodimeric scFv derivative. It consists of a fusion protein which contains the CH3 region of an immunoglobulin, preferably IgG, most preferably lgG1 as the dimerisation region which is connected to the scFv via a Hinge region (e.g. also from lgG1) and a linker region. Examples of minibody-antibody proteins are known from the prior art.

[0108]

[0106] By triabody the skilled person means a: trivalent homotrimeric scFv derivative. ScFv derivatives wherein VH-VL is fused directly without a linker sequence lead to the formation of trimers.

[0109]

[0107] The skilled person will also be familiar with so-called miniantibodies which have a bi-, tri- or tetravalent structure and are derived from scFv. The multimerisation is carried out by di-, tri- or tetrameric coiled coil structures. In a preferred embodiment of the present invention, the gene of interest is encoded for any of those desired polypeptides mentioned above, preferably for a monoclonal antibody, a derivative or fragment thereof.

[0110]

[0108] Further encompassed is a single-domain antibody (sdAb), also be referred to as nanobody, which is an antibody fragment of a single monomeric variable antibody domain. Single-domain antibodies are typically engineered from heavy chain antibodies found in camelids (VHH fragments) or cartilaginous fishes (VNAR fragments).

[0111]

[0109] The immunoglobulin fragments composed of the CH2 and CH3 domains of the antibody heavy chain are called “Fc fragments”, “Fc region” or “Fc” because of their crystallization propensity (Fc = fragment crystallizable). These may be formed by protease digestion, e.g. with papain or pepsin from conventional antibodies but may also be produced by genetic engineering. The N-terminal part of the Fc fragment might vary depending on how many amino acids of the hinge region are still present.

[0110] Antibodies comprising an antigen-binding fragment and an Fc region may also be referred to as full-length antibody. Full-length antibody may be monospecific and multispecific antibodies. Multispecific antibodies are antibodies which have at least two different antigen-binding sites each of which bind to different epitopes. A multispecific antibody includes, without being limited thereto, bispecific and trispecific antibodies. A bispecific antibody has two different binding sites. Multispecific antibodies also include antibody formats other than full-length antibodies such as antibody-derived molecules.

[0112]

[0111] Bispecific antibodies typically combine antigen-binding specificities for target cells (e.g., malignant B cells) and effector cells (e.g., T cells, NK cells or macrophages) in one molecule. Exemplary bispecific antibodies, without being limited thereto are symmetric antibodies with an additional scFv fused at the Fc part for second target, diabodies, BiTE (Bispecific T-cell Engager) formats and DART (Dual-Affinity Re-Targeting) formats. The diabody format separates cognate variable domains of heavy and light chains of the two antigen binding specificities on two separate polypeptide chains, with the two polypeptide chains being associated non-covalently. The DART format is based on the diabody format, but it provides additional stabilization through a C-terminal disulfide bridge. Trispecific antibodies are monoclonal antibodies which combine three antigen-binding specificities. They may be built on bispecific-antibody technology that reconfigures the antigen-recognition domain of two different antibodies into one bispecific molecule or may be based on a DART architecture with an scFv at the Fc part of one chain. For example, trispecific antibodies have been generated that target CD38 on cancer cells and CD3 and CD28 on T cells. Multispecific (bi- and tri-specific) antibodies are particularly difficult to produce with high product quality.

[0113]

[0112] The term “antibody-derived molecule” as used herein refers to any molecule comprising at least an antigen-binding moiety that is structurally related to antibodies. It includes bispecific antibodies and modified full-length mono- or bispecific antibodies, optionally further modified with an additional antigen binding moiety or smaller antibody formats, including the ones described herein. These antibody-derived molecules may also be referred to as antibody-like molecules.

[0114]

[0113] The term “antibody mimetic” as used herein refers to proteins that bind to specific antigens in a manner similar to antibodies, but that are not structurally related to antibodies. Antibody mimetic includes, without being limited thereto an anticalin, an affibody, an adnectin, a monobody, a DARPin, an affimer, and an affitin.

[0115]

[0114] A single-domain antibody (sdAb) may also be referred to as nanobody. The person skilled in the art will understand that the protein may comprise more than one antigen-binding domain and hence may be multivalent, preferably bivalent (e.g., a bivalent sdAb or a bivalent anticalin or any other bivalent antibody mimetic).

[0115] It was shown that the inherent properties of monoclonal antibodies (mAbs) require a tailored purification process which, even when highly optimized, is still unlikely to completely deplete all conceivable HCPs (Maier, M. et al., Biotechnology and bioengineering, 2024, 121 (10): 3181-3195; Oh, Y.H. et al., Biotechnology and bioengineering, 2024, 121(4), 1284- 1297). Thus, the CHO cells according to the invention may be particularly suitable and useful for the production of certain antibodies, particularly antibodies comprising unacceptable hydrolase activity in the composition following (optimized) purification. Traceless and simultaneous removal of several hydrolytic HCPs by means of downstream processing is even more challenging for the more diverse antibody-derived molecules (including bispecific antibodies and modified full-length mono- or bispecific antibodies, optionally further modified with an additional antigen-binding moiety or smaller antibody formats) and / or single-chain formats, which may need adapted purification trains due to their diverse physiochemical properties. The present invention, and particularly the cells according to the present invention are therefore particularly suitable and useful for the production of antibodies and antibody- derived molecules and particularly antibody-derived molecules.

[0116]

[0116] Another preferred therapeutic protein is a fusion protein, such as an Fc-fusion protein. Thus, the invention can be advantageously used for production of fusion proteins, such as Fc-fusion proteins. The effector part of the fusion protein can be the complete sequence or any part of the sequence of a natural or modified heterologous protein. The immunoglobulin constant domain sequences may be obtained from any immunoglobulin subtypes, such as lgG1 , lgG2, lgG3, lgG4, lgA1 or lgA2 subtypes or classes such as IgA, IgE, IgD or IgM. Preferentially they are derived from human immunoglobulin, more preferred from human IgG and even more preferred from human lgG1 and lgG2. Non-limiting examples of Fc-fusion proteins are MCP1-Fc, ICAM-Fc, EPO-Fc and scFv fragments or the like coupled to the CH2 domain of the heavy chain immunoglobulin constant region comprising the N-linked glycosylation site. Fc-fusion proteins can be constructed by genetic engineering approaches by introducing the CH2 domain of the heavy chain immunoglobulin constant region comprising the N-linked glycosylation site into another expression construct comprising for example other immunoglobulin domains, enzymatically active protein portions, or effector domains. Thus, an Fc-fusion protein according to the present invention comprises also a single chain Fv fragment linked to the CH2 domain of the heavy chain immunoglobulin constant region comprising, e.g., the N-linked glycosylation site.

[0117]

[0117] The term “cytokine” refers to small proteins, which are released by cells and act as intercellular mediators, for example influencing the behavior of the cells surrounding the secreting cell. Cytokines may be secreted by immune cells or other cells, such as T-cells, B- cells, NK cells and macrophages. Cytokines may be involved in intercellular signaling events, such as autocrine signaling, paracrine signaling and endocrine signaling. They may mediate a range of biological processes including, but not limited to immunity, inflammation, and hematopoiesis. Cytokines may be chemokines, interferons, interleukins, lymphokines or tumor necrosis factors.

[0118]

[0118] As used herein, “growth factor” refers to proteins or polypeptides that are capable of stimulating cell growth.

[0119]

[0119] The methods according to the invention are in vitro methods of culturing CHO cells used for high expression of a product of interest, such as a heterologous / recombinant protein or a RNA products. Examples of CHO cells commonly used for recombinant protein production are CHO, CHO-K1 , CHO-DXB11 (also referred to as CHO-DUKX or DuxB11), CHO-S cells and CHO-DG44 cells or the derivatives / progenies of any of such cell line. Particularly preferred are CHO cells, such as CHO-DG44 and CHO-K1 cells. Most preferred are CHO-DG44 cells. Glutamine synthetase (GS)-deficient derivatives of the mammalian cell, particularly of the CHO-DG44 and CHO-K1 cell are also encompassed. These cells are particularly suitable for GS-based selection (such as methionine sulfoximine (MSX) selection) of clones stably expressing the heterologous protein. In one embodiment of the invention the mammalian cell is a Chinese hamster ovary (CHO) cell, preferably a CHO- DG44 cell, a CHO-K1 cell, a CHO DXB11 cell, a CHO-S cell, a CHO GS deficient cell or a derivative thereof. The CHO cell may further comprise one or more expression cassette(s) encoding a heterologous or recombinant protein, such as a therapeutic protein, preferably a recombinant secreted therapeutic protein. Exemplary CHO cells are also listed in Table 1 below.

[0120] Table A: Mammalian production cell lines

[0121]

[0120] CHO cells are most preferred, when being established, adapted, and completely cultivated under serum free conditions, and optionally in media, which are free of any protein / peptide of animal origin. Commercially available media such as Ham's F12 (Sigma, Deisenhofen, Germany), RPMI-1640 (Sigma), Dulbecco's Modified Eagle's Medium (DMEM; Sigma), Minimal Essential Medium (MEM; Sigma), Iscove's Modified Dulbecco's Medium (IMDM; Sigma), CD-CHO (Invitrogen, Carlsbad, CA), CHO-S-lnvitrogen), serum- free CHO Medium (Sigma), and protein-free CHO Medium (Sigma) are exemplary appropriate nutrient solutions. Any of the media may be supplemented as necessary with a variety of compounds, non-limiting examples of which are recombinant hormones and / or other recombinant growth factors (such as insulin, transferrin, epidermal growth factor, insulin like growth factor), salts (such as sodium chloride, calcium, magnesium, phosphate), buffers (such as HEPES), nucleosides (such as adenosine, thymidine), glutamine, glucose or other equivalent energy sources, antibiotics and trace elements. Any other necessary supplements may also be included at appropriate concentrations that would be known to those skilled in the art. For the growth and selection of genetically modified cells expressing a selectable gene a suitable selection agent is added to the culture medium.

[0122]

[0121] According to the methods of the present invention the therapeutic protein produced in the CHO cell according to the invention may be purified and optionally formulated. Thus, pharmaceutically acceptable excipients may be added to the purified therapeutic protein to form a pharmaceutical composition comprising said therapeutic protein. In certain embodiments the pharmaceutical composition comprises PS, such as PS20 or PS80. Thus, PS20 or PS80 and optionally further excipients may be added. In certain embodiments the composition comprises PS and less than about 10%, less than about 5% or less than about 2% or less than about 1% of PS is degraded when the composition is stored at about 2°C to about 8°C, preferably about 5°C, for at least six months, and / or less than about 15%, less than about 10%, less than about 5% or less than about 2 % of PS is degraded when stored at about 2°C to about 8°C, preferably about 5°C for at least 12 months. The temperature for determining storage stability as outlined above is 2°C to 8°C, preferably 4°C to 6°C, more preferably about 5°C. In one embodiment about 0.2 % (w / v) or more PS20 or PS80, more preferably about 0.4 % (w / v) or more PS20 or PS80, more preferably about 0.4 % (w / v) PS20 is added.

[0123]

[0122] A pharmaceutical composition comprising the purified therapeutic protein produced in the CHO cells according to the present invention has reduced hydrolase activity and / or reduced PS degrading activity. Preferably, pharmaceutical composition comprising the purified therapeutic protein produced in the CHO cells according to the present invention has reduced hydrolase activity and / or reduced PS degrading activity compared to a pharmaceutical composition (reference composition) comprising the purified therapeutic protein produced in a CHO cell comprising (i) a lipoprotein lipase (LPL) gene knockout, but not comprising (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters (each) comprising at least one carboxylesterase gene; or even more preferably not comprising (ii) at least three further hydrolase gene knockouts, and / or (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene. The person skilled in the art will understand that the reference composition comprises the same pharmaceutically acceptable excipients (i.e., the same formulation) as the pharmaceutical composition produced in the CHO cells according to the invention.

[0124]

[0123] The therapeutic antibody, such as an antibody is produced in CHO cells in cell culture according to the method of the invention. Following expression, the therapeutic protein is harvested, further purified and optionally formulated. The therapeutic protein, such as an antibody, may be recovered from the culture medium as a secreted protein in the HCCF or from a cell lysate (i.e., the fluid containing the content of a cell lysed by any means, including without being limited thereto enzymatic, chemical, osmotic, mechanical and / or physical disruption of the cell membrane and optionally cell wall) and purified using techniques described herein. According to the invention the method comprises harvesting the recombinant protein of interest in the HCCF. Preferably the recombinant protein of interest (e.g., an antibody) is recovered (i.e. purified) from the HCCF following cell separation, such as by filtration and / or centrifugation. Thus, in certain embodiments the harvest includes centrifugation and / or filtration to produce a HCCF. Methods for purification are well known in the art and include chromatographic methods, such as affinity chromatography, ion exchange chromatography and / or hydrophobic interaction chromatography. Moreover, filtration such as membrane and / or depth filtration may be applied.

[0125]

[0124] In yet another aspect the invention relates to a composition comprising a therapeutic protein, wherein the therapeutic protein is obtainable by the methods according to the invention. In certain preferred embodiments the composition is a pharmaceutical composition further comprising pharmaceutically acceptable excipients. Preferably the pharmaceutical composition further comprises PS (PS20 or PS80), preferably at a concentration of 0.1 g / L or more. In certain embodiments, the pharmaceutical composition comprises PS and less than about 10% of PS is degraded when the composition is stored at about 2°C to about 8°C, preferably about 5°C, for at least six months.

[0126]

[0125] The pharmaceutical composition comprising the therapeutic protein according to the invention has reduced hydrolase activity and / or reduced PS degrading activity. Preferably the pharmaceutical composition comprising the therapeutic protein has reduced hydrolase activity and / or reduced PS degrading activity compared to a pharmaceutical composition (reference composition) comprising the purified therapeutic protein produced in a CHO cell comprising (i) a lipoprotein lipase (LPL) gene knockout, but not comprising (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene cluster (each) comprising at least one carboxylesterase gene; or even more preferably not comprising (ii) at least three further hydrolase gene knockouts, and / or (iii) two or more genome excisions of a gene cluster comprising at least one carboxylesterase gene.

[0127]

[0126] In view of the above, it will be appreciated that the invention also encompasses the following items:

[0128]

[0127] Item 1 provides a CHO cell comprising, (i) a lipoprotein lipase (LPL) gene knockout, (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) of the CHO genome, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C).

[0129]

[0128] Item 2 further specifies the CHO cell of item 1 , wherein at least one of the two or more genome excisions removes 1000 kb or more, preferably 1500 kb or more of the CHO genome.

[0130]

[0129] Item 3 further specifies the CHO cell of item 1 or 2, wherein the LPL knockout, the at least three hydrolase knockouts and the two or more genome excisions are biallelic.

[0131]

[0130] Item 4 further specifies the CHO cell of any one of items 1 to 3, wherein the two or more genome excisions are genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C).

[0132]

[0131] Item 5 further specifies the CHO cell of item 4, wherein the two or more genome excisions are genome excisions (a) of the CES1 cluster removing at least 1000 kb and (b) of the CES2 cluster removing of at least 500 kb of the CHO genome.

[0133]

[0132] Item 6 further specifies the CHO cell of any one of items 1 to 5, wherein the two or more genome excisions deleting at least two gene clusters comprising a carboxylesterase gene additionally comprises upstream and / or downstream flanking genomic regions, wherein the upstream and / or downstream flanking genomic regions are preferably 200 kb, 300 kb or 600 kb or more.

[0133] Item 7 further specifies the CHO cell of any one of items 1 to 6, wherein the CHO cell comprises at least one additional genome excision of a gene cluster and / or of at least 800 kb of the CHO cell genome.

[0134]

[0134] Item 8 further specifies the CHO cell of any one of items 1 to 7, wherein the CHO cell further comprises a glutamine synthetase (GS) knockout.

[0135]

[0135] Item 9 further specifies the CHO cell of any one of items 1 to 8, wherein gene knockout comprises gene excision and gene inactivation, preferably gene excision.

[0136]

[0136] Item 10 further specifies the CHO cell of any one of items 1 to 9, wherein the (ii) at least three further hydrolase gene knockouts comprise at least three knockouts of genes coding for hydrolases (a) selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), plateletactivating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2), and / or (b) of the hydrolase class with the EC number 3.1 and comprising a catalytical triad consisting of a serine (Ser)-histidine (His)-aspartate / glutamate (Asp / Glu) motif.

[0137]

[0137] Item 11 further specifies the CHO cell of any one of items 1 to 10, wherein the (ii) at least three further hydrolase gene knockouts comprise knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2).

[0138]

[0138] Item 12 further specifies the CHO cell of item 11 , wherein the (ii) at least three further hydrolase gene knockouts comprise knockouts of hydrolase genes selected from the group consisting of lipase A (LIPA), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2).

[0139]

[0139] Item 13 further specifies the CHO cell of any one of the preceding items, wherein the CHO cell comprises, (i) a lipoprotein lipase (LPL) gene knockout (ii) at least five further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deletes at least two gene clusters comprising at least one carboxylesterase gene.

[0140]

[0140] Item 14 further specifies the CHO cell of item 13, wherein the (ii) at least five further hydrolase gene knockouts comprise at least three knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH 1 ), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2).

[0141]

[0141] Item 15 further specifies the CHO cell of item 14, wherein the (ii) at least five further hydrolase gene knockouts comprise knockouts of genes coding for hydrolases selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2).

[0142]

[0142] Item 16 further specifies the CHO cell of any one of items 1 to 15, wherein the CHO cell further comprises a gene knockout of the gene coding for alpha-(1 ,6)-fucosyltransferase (FUT8).

[0143]

[0143] Item 17 further specifies the CHO cell of any one of the preceding items, wherein the CHO cell further comprises a Bax and Bak1 gene knockout.

[0144]

[0144] Item 18 further specifies the CHO cell of any one of items 1 to 17, wherein the CHO cell further comprises a recombinant gene encoding a therapeutic protein.

[0145]

[0145] Item 19 provides a method of producing a recombinant protein, comprising (a) introducing a nucleic acid comprising a gene coding for a therapeutic protein and a gene coding for a selection marker into the CHO cell of any one of items 1 to 17; (b) culturing the cell of step (a) under conditions to produce the therapeutic protein; (c) harvesting the therapeutic protein; and (d) purifying and optionally formulating the therapeutic protein.

[0146]

[0146] Item 20 provides a pharmaceutical composition comprising a recombinant protein produced according to the method of item 19.

[0147]

[0147] Item 21 provides a method of reducing polysorbate degrading activity in a CHO cell, comprising (a) providing a CHO cell; (b) knocking out (i) lipoprotein lipase (LPL) gene and (ii) at least three further hydrolase gene knockouts in the CHO cell; and (c) excising genomic regions comprising (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) in the CHO cell, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C).

[0148]

[0148] Item 22 provides a method of reducing polysorbate degrading activity in a recombinant protein produced in a CHO cell, wherein the method comprises (a) providing a CHO cell; (b) knocking out (i) lipoprotein lipase (LPL) gene and (ii) at least three further hydrolase gene knockouts in the CHO cell; (c) excising genomic regions comprising (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) in the CHO cell, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C); (d) transfecting the CHO cell comprising the knockout of step (b) and the two or more genome excisions of step (c) with a gene coding for a recombinant protein, preferably a therapeutic protein, and a gene coding for a selection marker; (e) harvesting the therapeutic protein; and (f) purifying and optionally formulating the therapeutic protein.

[0149]

[0149] Item 23 further specifies the method of item 21 or 22, wherein the CHO cell is the CHO cell according to items 1-17.

[0150]

[0150] Item 24 provides a use of the cell of item 18 for the production of a therapeutic protein.

[0151] EXAMPLES

[0152] Material and Methods

[0153] Hydrolase identification by LC / MS-MS

[0154]

[0151] Sample preparation: For each mAb, 8 mg of Ultrafiltration / Diafiltration material was diluted using purified water to the final volume of 370 pl, and pH was adjusted using 10 pL of 1 M Tris / HCI buffer (pH 8.0). 4 pg of trypsin was added to each sample, and samples were incubated at 37 °C for 2 h, at 550 RPM. Dithiothreitol was added to each sample to 13.4 mM final concentration, followed by 10 min incubation at 95 °C for 10 min, 550 RPM. Undigested mAb was pelleted by 5 min centrifugation at 15000 g, and supernatants were transferred to fresh tubes. Samples were acidified using 15 pL of 10% formic acid and desalted using C18 StageTips and dried by vacuum centrifuge. Purified dried peptides were fractionated using tip-based basic reversed-phase (BRP) fractionation. For BRP- fractionation 10 mM ammonium formate elution buffers were prepared with ascending ACN concentration. The pH of the elution buffers was adjusted to 10 using 28% NH4OH and BRP tip columns were prepared using 14 layers of C18 Empore Extraction material (3M). Columns were equilibrated by applying 100 pl of MeOH, 100 pl of 10 mM ammonium formate, 20% ACN, and 100 pl of 10 mM ammonium formate, 5% ACN. Dried samples were resolubilized in 200 pl of 10 mM ammonium formate, 5% ACN and were loaded to columns, and flow- through was collected as the first fraction after 3 min centrifugation at 2500 g. Subsequent fractions were collected using the elution buffers with 7.5%, 10.0%, 12.5%, 15.0%, 17.5%, 20.0%, and 50.0% ACN, and all fractions were dried using a vacuum centrifuge until further analysis.

[0155]

[0152] LC-MS / MS Data Acquisition: The dried samples were reconstituted in a solution containing 5% acetonitrile (MeCN) and 5% formic acid (FA), and then loaded onto a PepMap® RSLC C18 Easy-spray column (#ES903, ThermoFisher, Waltham, MA, USA) at a flow rate of 300 nL / min using solvent A (0.1 % FA in water) at a column temperature of 60 °C, using a Dionex UltiMate 3000 RSLCnano system (ThermoFisher Scientific, Waltham, MA, USA). Peptides were separated with a 240-minute linear gradient, starting from 98% solvent A and 2% solvent B (0.1% FA in MeCN) to 65% solvent A and 35% solvent B, at a flow rate of 250 nL / min. The column was washed with four cycles of zigzag washing steps, ranging from 2% solvent B to 95% solvent B.

[0156]

[0153] The eluted peptides from the analytical column were subjected to positive ionization at 2.0 kV using the EASY-Spray™ Source (ThermoFisher Scientific, Waltham, MA, USA) of an Exploris 480 mass spectrometer, equipped with a FAIMS pro interface (ThermoFisher Scientific, Waltham, MA, USA). The mass spectrometer was operated in DDA mode with survey scans acquired from m / z 300 to 1600 in the Orbitrap analyzer at a resolution of 60000 at four different compensation voltages (CV -40, -65, -75, -90), followed by fragmentation of the most abundant ions. MS / MS spectra were obtained using higher-energy collisional dissociation (HCD) at 30%. The isolation window was set at 2 m / z, the orbitrap resolution at 15000, the target value at 1 E5, and the maximum injection time set at auto. Selected precursor ions for fragmentation (including charge state 2-8) were excluded for 30 s, and the repeat count was set at 1.

[0157]

[0154] Data Analysis: Acquired raw files were analyzed using Proteome Discoverer 2.2 with the Mascot and SEQUEST algorithms in one combined search. The search was performed against the Chinese hamster UniProt proteome database (UP000001075). The applied search parameters were as follows: a full tryptic digestion and a maximum missed cleavage of 3, dynamic modifications of oxidation at methionine and acetylation at protein N-termini, the minimum peptide length of 6, and a maximum of 144. The precursor mass tolerance and fragment mass tolerance were set to 10 ppm and 0.02 Da, respectively. Data filter criteria were set as 1% and 5% FDR for peptides and proteins, respectively. A protein was considered as identified when at least two unique peptides fulfilling the mentioned criteria could be identified. Uniprot51 and BRENDA52 databases were used for the enzyme class determination of the HCPs.

[0158] Hydrolase quantification by LC / MS-MS

[0159]

[0155] Sample preparation: An equivalent of 100pg antibody from harvested cell culture supernatant were denaturized with 80 pl of 7 M guanidine- HCI, 0.4 M tris-HCI buffer at a final volume of 180 pl. Samples were reduced in 5mM DTT for 30min at 56°C for 30 min and alkylated with 10mM Chloroacetamide at room temperature in the dark for 45 min. Afterwards, samples were diluted to a final volume of 800 pl with 25 mM Tris-HCI before adding 1.5 pg of Trypsin / LysC mix (#V5073, Promega, Madison, Wl, USA). Digestion was carried out overnight for at least 18h. Digested samples were desalted using Pierce™ Peptide Desalting Spin Columns (#89851 , ThermoFisher Scientific, Waltham, MA, USA) as described in the manufacturer’s manual and dried until further use.

[0156] LC-MS / MS Data Acquisition: The dried samples were reconstituted in 2% acetonitrile (MeCN) and 0.1% formic acid (FA) and trapped on a PepMap™ Neo 5 pm C18 300 pm x 5 mm Trap cartridge (#174500, ThermoFisher Scientific, Waltham, MA, USA) at a flow rate of 30 pl / min using a Dionex UltiMate 3000 RSLCnano System (ThermoFisher Scientific, Waltham, MA, USA). Subsequently, peptides were eluted in back flush on to a PepMap® RSLC C18 Easy-spray column (#ES75500PN, ThermoFisher Scientific, Waltham, MA, USA) (75 mm inner diameter x 50 cm, 2 mm particle size, 100 A) with a 120-minute linear gradient at a flow rate of 300nL / min, starting from 98% solvent A and 2% solvent B (0.1 % FA in MeCN) to 65% solvent A and 35% solvent B, at a flow rate of 300 nl / min. Afterwards the column was washed with 80% solvent B before being equilibrated with 98% solvent A and 2% solvent B.

[0160]

[0157] The eluted peptides were subjected to ionization at 2.0 kV using an EASY-Spray™ Source (ThermoFisher Scientific, Waltham, MA, USA) of an Exploris 480 mass spectrometer, equipped with a FAIMS pro interface (ThermoFisher Scientific, Waltham, MA, USA). The mass spectrometer was operated in data dependent acquisition (DDA) mode with full scans acquired from m / z 300 to 1600 in the orbitrap analyzed at a resolution of 60,000 followed by fragmentation of the 10 most abundant ions at three different compensation voltages (CV - 50, -65, -85). Tandem mass spectra were obtained using higher-energy collisional dissociation at 30%. The isolation window was set at 2 m / z, the Orbitrap resolution at 15,000, the target value at 1 E5, and the maximum injection time set at auto. Selected precursor ions for fragmentation (including charge state 2-8) were excluded for 45s, and the repeat count was set at 1.

[0161]

[0158] Data Analysis: Raw files were analyzed using Proteome Discoverer 3.0 using Mascot and SEQUEST search algorithms in one combined search using the Chinese Hamster TrEMBL proteome database (downloaded 01 .03.2023, 57854 entries). Database search was carried out using search parameters as follows: full tryptic digestion and a maximum missed cleavage of 2, dynamic modifications of oxidation at methionine and acetylation at protein N- termini, the minimum peptide length of 6, and a maximum of 144. The precursor mass tolerance and fragment mass tolerance were set to 10 ppm and 0.02 Da, respectively. Data filter criteria were set as 1% false discovery rate (FDR) for peptides and proteins. Peak areas deriving from the full scans were used for relative quantification using the precursor ions quantifier node with default settings in Proteome Discoverer.

[0162] Cell culture

[0163]

[0159] CHO K1 producer cell lines were cultured in 125 mL shake flasks on an orbital shaker (Infors HT, Bottmingen, Switzerland) with 120 rpm (50-mm orbit). Cells were seeded with 5 x 105viable cells / mL and passaged every 2-3 days. Viable cell density (VCD) and viability were determined by trypan blue exclusion using Vi-CELL BLU (Beckman Coulter, Brea, CA, USA). Cultured cells were maintained at 36.5 °C and 5 % CO2.

[0164] Generation of LPL KO cell lines

[0165]

[0160] Generation of LPL KO cell lines producing mAb was done via CRISPR / Cas9 deletion technology. In brief, 180 pmol Alt-R™ Sp. Cas9 (#1081059, Integrated DNA Technologies) was complexed to 220 pmol Alt-R™ sgRNAs (Integrated DNA Technologies) (Upstream: TCTACCGCGCTCCAGCCCTT (SEQ ID NO: 33); Downstream:

[0166] GGTCTTGGCGCACTCTAGAA (SEQ ID NO: 34)) and transfected via the NEON electroporation system (Thermo Fisher Scientific, Waltham) in 5x106cells. Transfected cells were resuspended in 25 mL culture medium in 125 mL shaking flasks and single cell sorted via FACSAriaTM (BD Bioscience, San Jose) after 3 days cultivation. Following expansion, biallelic LPL KO clones were identified via (Non-)Deletion PCR of extracted gDNA (NonDeletion primerfw: tggggcaacggtactgtaga (SEQ ID NO: 35); rev: aacgtaccgtctgctgcg (SEQ ID NO: 36); Deletion primerfw: tggggcaacggtactgtaga (SEQ ID NO: 37); rev: ctggatgtgaggaagctacttcc (SEQ ID NO: 38)). In addition, lack of LPL expression on protein level was confirmed by ELISA measurement (#SEA386Mu, Cloud-Clone Corp.).

[0167] Generation of stably expressing hydrolase cell lines and recombinant hydrolase expression

[0168]

[0161] For hydrolase expression, a transposase system was used in combination with a glutamine selection system. Hydrolase sequences were extracted from the proprietary CHO K1 transcriptome and cloned into a transposase recognition sequence harboring expression plasmid. For purification purposes, a C-terminal His- and Strep Tag were added to the native hydrolase coding sequences. 3 pg transposon plasmid and 1.5 pg transposase expression plasmid were transfected (NEON electroporation system) into 5x106cells and cells were resuspended in 5 mL medium and incubated statically. Selection pressure was applied one day post-transfection and cells were expanded in 40 mL medium. After reaching a viability of >40% and a VCC of >1.5x105cells / mL, cells were transferred in 25 mL medium in a shaking flask and cultivated until reaching viabilities >95%. Stable hydrolase expressing cell pools were (fed-)batch cultivated in 1.5 L cultures over 5-12 days and cell pellet / supernatant was harvested via centrifugation and stored at -70°C.

[0169] Enzyme localization

[0170]

[0162] Expression and localization of the protein of interest was analyzed with biolayer interferometry using Anti-His2 biosensors (Octet® Anti-HIS2 #18-5114, Sartorius, Gottingen, Germany) on an Octet HTX platform (Sartorius, Gottingen, Germany). For protein localization, cell pellets were lysed using the same volume of lysis buffer (50 mM Tris pH 8, 150 mM NaCI, 1 % Triton X-100) as the corresponding cell culture supernatant. This ensured a consistent comparison between intracellular and extracellular protein levels. The ratio of the amount of intracellular- and extracellular protein was then calculated.

[0171] Purification of recombinant hydrolases

[0172]

[0163] Recombinant hydrolases were purified using two orthogonal capture steps followed by size exclusion chromatography on an Akta Avant 150 system (Cytiva, Marlborough, MA, USA). Initially, immobilized metal affinity chromatography (IMAC) was performed using a nickel-nitrolotriacetic acid column (HisTrap FF crude #17528601 , Cytiva, Marlborough, MA, USA). The column was equilibrated with buffer containing 20 mM Hepes pH 8, 150 mM NaCI and 20 mM imidazole. After sample loading, the column was washed with the same buffer until baseline absorbance at 280 nm. The bound protein was eluted by an imidazole gradient over 10 column volumes (20-500 mM). Elution fractions were pooled and loaded onto a Strep-Tactin column (StrepTrap XT #29401323, Cytiva, Marlborough, MA, USA) according to manufacturer’s instructions. Proteins were eluted with a step elution of 10 mM HEPES, 150 mM NaCI and 50 mM biotin at pH 8. Samples containing the protein of interest, as determined by SDS-PAGE, were pooled and polished by size exclusion chromatography (HiLoad 16 / 600 Superdex 200 pg #28989335, Cytiva, Marlborough, MA, USA). The recombinant hydrolase monomers were identified by comparing the chromatogram to a molecular weight standard. Fractions containing pure protein were aliquoted and stored at - 70 °C. pH dependent activity

[0173]

[0164] The esterase activity assay was fully automated using a Fluent Control (Tecan Group, Mannerdorf, Switzerland) and employed 4-methylumbelliferyl decanoate (MUD4) as a substrate to evaluate hydrolytic activity at various pH values between 4 and 8. Multicomponent buffers were prepared at pH 4 and 8, and a master mix containing the MUD4 substrate was created. The multi-component buffers were mixed in a 12-column reservoir trough to generate a linear pH gradient. The assay was performed in a 96-well plate with the following composition per well: 75 p sample, 75 pL water, 150 pL master mix at various pH values. MUD4 is a nonfluorescent ester that, upon cleavage by esterases, releases the fluorescent product 4-methylumbelliferon (MU). This cleavage of the ester bond results in an increase in fluorescence signal within the sample, which can be detected using excitation and emission wavelengths of 340 nm and 450 nm (Spectra Max M series, Molecular Devices, San Jose, CA, USA), respectively. By measuring the assay at different pH values, pH- dependent activities can be investigated. The fluorescence intensity is proportional to the amount of fluorogenic substrate (MUD4) being cleaved, allowing for the quantification of esterase activity in the sample under various pH conditions. Generation of 9x hydrolase KO cell lines

[0174]

[0165] A CHO-K1 cell line with a biallelic GS knockout (CHO-K1 GS) was used for the performed experiments, as well as CHO-K1 GS cell expressing different recombinant products. CHO cells were adapted for growth in chemically defined and animal component- free media in suspension culture.

[0175]

[0166] Deletion of LPL was performed with a zinc finger nuclease (ZFN) pair targeting exon 7 (Sigma-Aldrich). 2 pg of each ZFN were transfected with the NEON electroporation device (ThermoFisher, Waltham, MA, USA) in 1x106cells and transferred into 10 mL medium in a 50 mL spin tubes (TPP, Trasadingen, Switzerland). Single cell cloning was performed 3 days after transfection and LPL KO validation was executed via ELISA (Cloud-Clone Corp.)

[0176]

[0167] Deletion of all remaining hydrolases was performed via CRISPR / Cas9. 180 pmol Alt- R™ Sp. Cas9 (IDT) was complexed to 220 pmol Alt-R™ sgRNAs (IDT) and transfected via the NEON electroporation system in 5x106cells. Transfected cells were resuspended in 25 mL culture medium in 125 mL shaking flasks and single cell sorted after 3 days cultivation. After expansion, sanger sequencing of the Cas9 target site followed by interference of CRISPR edits (ICE) analysis was used to identify clones with biallelic out-of-frame insertion and / or deletions (InDeis). Biallelic gene excisions were identified via (non-) deletion PCR.

[0177]

[0168] The knock-out strategy is visualized in Figure 6 and the sgRNA used are listed in Table 4. The intermediate and final clones are summarized and described in Table 3.

[0178] Table 3: Overview of generated hydrolase knockout (KO) clones with respective genes knocked out.

[0179] Table 4: Overview of used sgRNA sequences to generate insertion and / or deletions

[0180] (InDeis) or gene excisions using CRISPR / Cas9. Lpl was knocked out using zinc finger nucleases.

[0181] RNA sequencing

[0182]

[0169] Total RNA was isolated from 5x106cells using QIAsymphony RNA Kit and the QIAsymphony (Qiagen, Hilden, Germany). The RNA quality was assessed using the Fragment Analyzer with the RNA Kit (Agilent, Santa Clara, CA, USA) and samples with RNA quality numbers of > 7.5 were processed further. Strand specific RNA libraries were prepared using the NEBNext Ultra II Directional RNA Library Kit and the NEBNext Poly(A) mRNA Magnetic Isolation Module (New England Biolabs, Ipswich, MA, USA). Libraries were sequenced on Illumina’s NovaSeq 6000 (Illumina, San Diego, CA, USA) in 100 bp paired- end mode generated mean read counts of 3 x 107read pairs per sample. Sequencing data was analyzed with GeneData Selector® software (Genedata, Basel, Switzerland). Sequence reads were quality filtered, adapter sequences were trimmed off and high-quality reads were mapped to CHO K1 genome. Protein coding sequences were annotated, counted and normalized to library size. Gene expression was calculated as transcripts per million reads (TPM). Generation of stably expressing monoclonal antibody (mAb) producer pools

[0183]

[0170] For mAb expression, a transposase system was used in combination with a glutamine selection system. 5 pg transposon mAb expression plasmid and 4 pg transposase mRNA were transfected (Maxcyte electroporation system, Rockville, MD, USA) into 5x106cells and cells were resuspended in 5 mL medium and incubated statically. Selection pressure was applied one day post-transfection and cells were expanded in 40 mL medium. After reaching a viability of >40% and a VCC of >1.5x105cells / mL, cells were transferred in 25 mL medium in a shaking flask and cultivated until reaching viabilities >95%.

[0184] Fed-batch cultivation

[0185]

[0171] Fed-batch cultivations were either performed using the ambr15® system (Sartorius, Gottingen, Germany) or manually in shaking flasks. Viable cell density and viability was measured daily via CEDEX (Roche Diagnostics, Rotkreuz, Switzerland) or Vi-CELL BLUE (Beckman Coulter, Brea, CA, USA). After seeding, feed was daily added from day 2 for ambr cultivations and day 3 for shaking flask runs. Glucose was added to 6 g / L if glucose concentration dropped below 4 g / L. ambr cultivations were cultivated for 14 days, shaking flask runs were cultivated for 11 days and harvested via centrifugation and 0.2pm sterile filtration. Titer was measured via biolayer interferometry employing the Octet® system (Sartorius) with protein A biosensors. Titer data was set against the integral of VCD to gain specific productivity (qp).

[0186] Purification of monoclonal antibody (mAb) fraction

[0187]

[0172] Harvested cell culture fluid (HCCF) from fed-batch was thawed and 0.2 pm filtered. MAb purification was performed via 0.6ml prismA robocolumns (Cytiva, Marlborough, MA, USA) with Tris and glycine containing wash and elution buffers, respectively.

[0188] HCP measurement

[0189]

[0173] HCCF and purified mAb fractions were measured via biolayer interferometry measurement using the Octet® system (Sartorius)in combination with anti-CHO HCP biosensors.

[0190] Hydrolytic activity assays

[0191]

[0174] MUD4 / MUN4 assay: Hydrolytic activity of HCCF and the purified mAb fraction was determined using 4-methylumbelliferyl decanoate (MUD4; FM25973, Carbosynth) or 4- methylumbelliferyl nonanoate (MUN4; EM52253, Carbosynth). MUD4 / MUN4 are nonfluorescent esters that, upon cleavage by esterases, releases the fluorescent product 4- methylumbelliferon. This cleavage of the ester bond results in an increase in fluorescence signal within the sample, which can be detected using excitation and emission wavelengths of 340 nm and 450 nm (MLID4) or 320 nm and 450 nm (MLIN4).

[0192]

[0175] The buffer used in the MUD4 / MUN4 assay is referred to as AMT buffer (or multicomponent buffer) with a broad buffering range of 4-8 provided as a 4x stock solution (0.3 M acetic acid, 0.3 M MES, 0.6 M TRIS, 0.6 M NaCI, 40 mM CHAPS) or 2x stock solution (0.15 M acetic acid, 0.15 M MES, 0.3 M TRIS, 0.3 M NaCI, 20 mM CHAPS), with the pH adjusted as indicated using HCI or NaOH, resulting in a final concentration in the reaction mixture of 75 mM acetate, 75 mM MES, 150 mM TRIS, 150 mM NaCI, and 10 mM CHAPS. Alternatively, a phosphate buffer as a 2x stock solution comprising 0.2 M phosphate buffer, 0.28 M NaCI, 20 mM CHAPS with a pH adjusted to pH 7.4 resulting in a final concentration in the reaction mixture of 0.1 M phosphate buffer, 0.14 M NaCI and 10 mM CHAPS may be used. For the MLIN4 assay the 2xATM buffer at pH 6 was used.

[0193]

[0176] The substrate is stored as a concentrated stock solution comprising 3 mM MLID4 or 3 mM MLIN4 in DMSO. MLID4 is diluted 1 :10 in DMSO prior to use resulting in a 100x stock solution for use comprising 0.3 mM in DMSO. MLIN4 is used undiluted resulting in a 100x stock solution for use comprising 3 mM in DMSO.

[0194]

[0177] The given sample (e.g. drug substance) is mixed with the master mix including the substrate and the assay buffer and optionally additional water. Per well in a 96 well plate mastermix (75 pL 4x stock assay buffer, 150 pL H2O, 3 pL 0.3 mM 4-MUD in DMSO or 150 pL 2x stock assay buffer, 75 pL H2O, 3 pL 0.3 mM 4-MUD or 3 mM 4-MUN in DMSO) has been added to 72 pL sample in the reaction vessel.

[0195]

[0178] Hydrolysis of the substrate MUD4 (or 4-MUD) / MUN4 (or MUN-4) has been measured by detecting the fluorescence intensity of the released chromophore 4-MU immediately following mixing in real-time for a few minutes up to 5 hours depending on fluorescence intensity, typically 90 min. Calibration curves with known 4-MU concentrations further allow the determination and comparison of reaction velocities at different pH values.

[0196]

[0179] Fluorescence Micelle Assay (FMA): The assay was performed as described before

[0197] (WeiB L, Schmieder-Todtenhaupt V, Haemmerling F, Lakatos D, Schulz P, Fischer S. Biotechnology and bioengineering 2023; PMID:37743807; htps: / / doi.Org / 10.1002 / bit.28563).

[0198]

[0180] For purified hydrolases the hydrolases were spiked into a typical formulation buffer (50 mM Histidine pH 6) at a final concentration of 25 pg / mL and mixture was incubated at room temperature. At different time points (dO, d1 , d3, d7, d14), samples were frozen at -70 °C. Hydrolase containing cell culture supernatant was diluted in PS spiked (0.4 mg / mL), fresh cultivation media.

[0181] The fluorescence micelle assay (FMA) was also employed to determine the polysorbate (PS) concentration in HCCF. PS20 was diluted in respective medium to obtain the PS-medium working solution. HCCF was mixed with the working solution to obtain PS spiked samples (cPS = 0.4 mg / ml). 200 pl was aliquoted in 5 tubes in a cooling rack. One tube, representing sample to, was immediately frozen at -70°C. Remaining tubes were incubated at RT in the dark and frozen at -70°C after 1 , 3, 7 and 14 days. Prior to measurement, standard curves ranging from 0.1 to 0.6 mg / ml were generated. For PS measurement, 10 pl sample was mixed with 240 pl FMA reagent (150 mM NaCI, 50 mM Tris, 0.2% Acetonitrile, 5 pM NPN, 0.0015% Brij-35, pH 8) and subsequently incubated for 1 min at 35°C. During incubation, fluorescent dye N-phenyl-1-naphtylamine intercalates in PS formed micelles. Fluorescent signal, positively correlating with PS concentration, was measured at 420 nm following excitation of the dye with a 350 nm laser using an Infinite® 200 PRO microplate reader (Tecan, Mannedorf, Switzerland).

[0199] Product quality (PQ) analysis

[0200]

[0182] Harvested supernatant from fed-batch was thawed and 0.2 pm filtered. MAb purification was performed via 0.6 ml prismA robocolumns (Cytiva) with wash (50 mM Tris, 500 mM Arginine, 1.0 M Urea, pH 7.5) and elution buffers (150 mM Glycine, pH 3.5) and product concentration was measured via absorption measurement at 280nm. High molecular weight (HMW) species in the purified mAb fraction was determined via size exclusion chromatography (SEC, 1290 Infinity II / Agilent Technologies) using a SEC column (Acquity UPLC BEH 200 SEC;1.7 pm; ID 4.6 mm x L 300 mm; 5 mL) with respective SEC buffer (200 mM Arginine, 120 mM Ammonium sulfate, 10% isopropanol, pH 7.3). Low molecular weight (LMW) species were measured via non-reduced capillary gel electrophoresis (cGE, LabChip / PerkinElmer). Charge heterogeneity was evaluated via a strong cation-exchange column (CEX, 1260 Infinity II / Agilent Technologies) to identify main, acidic (APGs) and basic (BPGs) IgG fractions. For lgG1 analysis CEX was performed on the column BioPro IEX SF, S-5pm, 100 x 4.6 mm (YMC) with wash (20 mM ACES pH 7.2) and elution buffer (20 mM ACES, 100 mM NaCI, pH 7.2) and for lgG4 analysis CEX was performed on a MabPac SCX- 10 10 pm / 4 x 250 mm 10 pm (ThermoScientific) with wash (20mM ACES, pH 6.2) and elution buffer (20mM ACES, 500mM NaCI, pH 6.2). N-glycosylation of the Fc part of the IgGs was analyzed by LabChip® glycan profiling assay (PerkinElmer).

[0201] Measurement of viable cell concentration and viability

[0202]

[0183] For the measurement of viable cell concentration (VCC) and viability of cells cultured in shaking flasks, the Vi-CELL BLUE (Beckman Coulter, Brea, CA, USA) was used. Per measurement, 200 pl cell suspension was pipetted in the designated sample vessel and fully automated measurement was initiated. The image-based technology makes use of the Trypan Blue exclusion method for cell counting. Dead cells with a damaged cell membrane will be stained by Trypan Blue, whereas live cells with an intact cell membrane exclude the dye (Strober 2001).

[0203] Additional Bax / Bak1 knockout

[0204]

[0184] In order to improve harvest viabilities Bax and Bak1 (BB), which are coding for the apoptosis-inducing proteins “apoptosis regulator BAX” and “Bcl-2 homologous antagonist / killer”, respectively, were simultaneously knocked-out. Because the multi-KO 2G11 cell line showed overall best performance during the production run, it was used as basis for the BB KO. Successful KO was verified via western blot demonstrating the lack of BAX and BAK protein bands in all 4 tested cell lines in comparison to the parental cell line (data not shown). Therefore, a 14-days fed-batch cultivation was performed, which confirmed the successful rescue of harvest viabilities from below 35% with multi-KO 2 G11 to approx. 90% for all BB KO cell lines on day 14 (data not shown).

[0205] Example 1 : Identification of hydrolytic Host Cell Proteins in formulated mAb products

[0206]

[0185] By means of a highly sensitive liquid chromatography-mass spectrometry (LC-MS / MS) technique (down to 0.3 ppm), 7 mAb products were analyzed for their residual hydrolase content. Of note, the analyzed mAb products were industrially relevant molecules that were purified through a chemistry, manufacturing, and controls (CMC) bioprocess and were formulated to simulate preparation for clinical application. Across all investigated products, 12 hydrolytic enzymes acting on ester bonds (enzyme class, EC 3.1) were identified (Table 5).

[0207] Table 5: List of hydrolytic host cell proteins of the enzyme class 3.1, identified at least once in LC / MS-MS analyses of 7 different antibody formulations.

[0208]

[0186] Although all mAb products were expressed using the same CHO expression platform the number of identified hydrolases in the purified antibody products varied strongly from 1 to 6 out of 7 investigated samples. Frequently identified hydrolases such as LPL, GNS and SMPD1 opposed hydrolases which were only found in single products (CES2C, CES1(F), PLD3, IAH1 , PAF-AH). To investigate whether the expression of identified hydrolases correlated with their frequency of occurrence in purified antibodies, we analyzed previously compiled transcriptomics and proteomics data sets of 3 different, representative cell line development campaigns (Figure 1 A+B). Overall, a high and statistically significant correlation of transcriptome and proteome data was observed for all 3 mAb datasets (data not shown). Strikingly, for LPL, a frequently mentioned hydrolase in the context of PS degradation the high occurrence in many mAb products correlated with high mRNA and protein expression in all investigated production cell lines. This was also true for SMPD1 and LIPA, although here high expression values correlated only with identification in 5 and 3 out of 7 analyzed mAb products, respectively. The hydrolases CES1 F and PAF-AH showed high gene and protein expression in one of the investigated mAb formulations. The other identified hydrolases were expressed at much lower levels and occurrence of identification was low, except for GNS, which was identified in 6 out of 7 products (Figure 1A). In conclusion, expression of hydrolases varies substantially in CHO cells and mRNA and protein expression levels are not sufficiently predictive of the occurrence and abundance of the hydrolase in mAb products, which is probably due to unique interactions of difficult-to-remove hydrolases with the product and / or the chromatography columns.

[0209] Example 2: Knockout of PS degrading lipoprotein lipase (LPL) in CHO cells to allow for unbiased HCP characterizations

[0210]

[0187] As LPL was most frequently identified in mAb product formulations, expressed at the highest level in CHO production cells and previously suggested as a source for substantial PS degradation, the extent of LPL-related PS degradation was investigated. Therefore, LPL was knocked out in a mAb-producing CHO cell line using CRISPR / Cas9 genome editing technology. A combination of two different polymerase chain reaction (PCR) experiments validated the complete bi-allelic removal of the LPL gene in the engineered production cell line (Figure 2A). The KO production cell line was fed-batch cultivated head-to-head with the respective originator mAb production cell line (expressing LPL) over a period of 14 days in an ambr®250 bioreactor system. Supernatant sampled on day three of the fed-batch process was measured using LPL enzyme-linked immunosorbent assay (ELISA), which confirmed the absence of LPL protein in samples of the LPL KO cell line (Figure 2B). Next, supernatant harvested at day 14 of the cultivation process was subjected to a fluorescence micelle assay (FMA) measuring PS20 degradation over time at five different time points post PS spiking (dO, d1 , d3, d7, d14) (Figure 2C). Strikingly, in the supernatant of the non-edited parental CHO production clone (LPL wildtype (wt)) 95% of initially spiked PS content was degraded already after one day of incubation. In contrast, supernatant of LPL KO production cells elicited a significantly slower PS degradation rate resulting in only 7% degradation of initial PS levels after one day of incubation and 31% degradation after 14 days. While these data suggest a major PS degrading activity by LPL, the observed residual PS degrading activity clearly points in the direction of further hydrolases to contribute to the total PS degradation activity. Therefore, an LPL KO cell line would certainly serve as an improved manufacturing host cell line for biologies but also represents an ideal host cell line to characterize the additionally present hydrolases without the substantially masking PS degradation activity of LPL. Example 3: Secretion analysis of CHO hydrolases

[0211]

[0188] In order to characterize the additional hydrolases previously identified in mAb products and to analyze their potential PS degradation activity (Table 5), the remaining 11 CHO hydrolases were individually overexpressed in a CHO K1 glutamine synthetase (GS) LPL KO host cell line to avoid the masking potential of LPL. Here, the KO was realized via zink finger nucleases (ZFN) and validated using LPL ELISA (Figure 3A). Native CHO coding sequences of the 11 hydrolases, containing no additional signal sequence, were cloned into a transposon expression plasmid including His- and Strep-Tags at the 3’-end of the hydrolase gene (Figure 3B). Next, transposon plasmids were stably introduced in the CHO K1 GS LPL KO host cell line using a transposase technology in combination with a GS based metabolic selection approach to generate stable hydrolase expressing pools. As many of the identified hydrolases were annotated by UniProt as intracellular proteins, the question was raised why the potentially intracellular hydrolases were identified in the harvested supernatant and ultimately in the formulated mAb products. To determine whether the hydrolases are actively secreted into the supernatant, their localization in supernatant and cell pellet was quantified via the His-Tag after a three-day batch cultivation of all generated hydrolase overexpressing cell pools that displayed > 98% viable cells at harvest (data not shown). While the hydrolases PAF-AH and SMPD1 were fully secreted by the CHO cells, other proteins including CES1 F, CES2C, GNS, LPLA2, and PPT1 were partially detected in the intracellular fraction. Only LIPA, IAH1 , and PLD3 were mainly localized intracellularly (Figure 3C). Other than PLD3, LIPA and IAH1 could be quantified in supernatant of the 3 mAb producer cell lines (Figure 1 B), which might be due to mAb-HCP shuttling mechanisms or simply a reduced harvest viability of the assayed cell culture (59, 84, 88% for mAb1 , 2, 3, respectively). As most of the overexpressed native CHO hydrolases are actively secreted their occurrence in mAb products is not surprising.

[0212] Example 4: Hydrolytic activity of CHO hydrolases against polysorbate

[0213]

[0189] A detailed characterization of all overexpressed CHO hydrolases regarding their PS degradation potential was performed. All stable hydrolase-expressing cell pools were cultivated in (fed-) batch mode and supernatants were harvested. The crude supernatant was spiked with PS20 (0.4 mg / mL) and PS degradation was investigated over a period of 14 days via FMA. Strong PS degradation activity was measured for hydrolases CES1 F, CES2C, PPT1 , and PAF-AH (Figure 4A). Furthermore, moderate PS degradation was elicited by LPLA2, and IAH1 , while no PS degradation was observed in supernatants of cell pools overexpressing GNS, SMPD1 , LIPA, CES1 and PLD3, which also served as a comparative baseline. Notably, the crude culture supernatant does not allow for the determination of the exact PS degradation activity for each enzyme, as the supernatant likely contained additional or lacked hydrolases that are not efficiently secreted. Thus, for a direct comparison with uniform starting concentrations a multi-step purification strategy was developed to obtain highly purified protein. Towards this end, overexpressed proteins were purified from harvested cell culture supernatant or cell pellets using a combination of two orthogonal capture steps including an immobilized metal ion affinity column (IMAC) specifically binding the His-Tag on the recombinant hydrolases followed by a Strep-Tactin column to further capture the Strep-tagged hydrolases. In addition, size exclusion chromatography was performed as a polishing step and for aggregate removal. After purification, the CHO hydrolases were tested for their catalytic activity towards PS20 via FMA. Contrary to the FMA performed with crude supernatant, purified enzymes were spiked at uniform concentrations in a formulation buffer containing PS20 and PS degradation was investigated over a period of 14 days as described before (Figure 4B). In contrast to the previous analysis of crude cell culture supernatant, purified CES1 , IAH1 and LIPA now showed strong PS degradation activity. This observation aligns with the fact that these enzymes are not secreted and, as a result, are not present in the crude supernatant (see Figure 3C). Strikingly, LPLA2 and PAF- AH now displayed the strongest PS degradation activity among all assayed hydrolases. For CES1 F and CES2C, similar PS degradation was observed, while the hydrolase PPT1 now exhibited lower PS degradation activity as a purified enzyme. As before, GNS, SMPD1 and PLD3 displayed no PS degradation potential. Besides the above-mentioned drawbacks when measuring crude cell culture supernatant, it is important to note that the stability or aggregation of purified enzymes can influence the observed activity levels over time. In comparison to other hydrolases, PPT1 was particularly prone to aggregation during the purification process (data not shown).

[0214]

[0190] These experiments showed that hydrolase containing crude cell culture supernatant can provide an initial indication of hydrolase specific PS degradation potential, however, purification of specific hydrolases is necessary to determine their specific enzymatic activity. In summary, 8 out of 11 hydrolases previously identified in the mAb formulations displayed significant PS degradation activity. Of note, enzyme activity was measured under uniform pH conditions, whereby individual pH optima of the enzymes were not considered.

[0215] Example 5: Comparison of active sites in the CHO hydrolases

[0216]

[0191] To identify common features among the identified active hydrolases, we compared the active sites of both PS degrading and non-degrading hydrolases. By referring to UniProt annotations we determined the active site residues of the CHO hydrolases. In addition, we compared the protein sequences to orthologous enzymes with experimentally determined active sites (Table 6). A common catalytic triad consisting of a serine (Ser) - histidine (His) - aspartate / glutamate (Asp / Glu) motif was identified in PS degrading hydrolases, which distinguished them from non-PS degrading enzymes (Table 6).

[0217] Table 6: Active site residues of hydrolases identified in 7 industrial mAb formulations. Residues based on indicated literature and UniProt.

[0218] *residue positions are not known

[0219]

[0192] Enzymes with alternative catalytic sites (GNS, PLD3 and SMPD1) did not exhibit PS degradation activity. This finding is particularly valuable when new hydrolases are identified, as it provides a clear marker of potential PS degradation activity. It is important to note that previous research has shown that PS-degrading hydrolases are not limited to the enzyme sub-class of “carboxylic ester hydrolases (3.1.1). In fact, they can also be found within the broader enzyme class “acting on ester bonds” (3.1), which includes not only carboxylic-ester hydrolases but also thioester and phosphoester hydrolases, among others. This underscores the diverse enzymatic landscape involved in PS degradation and highlights the importance of the conserved catalytic triad as a predictive marker for PS degradation activity.

[0220] Example 6: pH dependent activity of hydrolytic HCPs

[0221]

[0193] The activity optimum of enzymes is strongly dependent on an ideal pH.31 To further characterize and evaluate the pH optima of each identified hydrolase acting on PS, pH screenings with the purified enzymes were conducted (Figure 5A). Here, enzymatic activity was measured upon cleavage of ester bonds in an assay employing 4-methylumbelliferyl decanoate (MLID4) as a PS surrogate. LIPA and LPLA2 displayed unique activity profiles with activity optima in a neutral or acidic pH, respectively. In contrast, the group of CES1 , CES2C, IAH1 and PAF-AH revealed similar pH dependent activity patterns with low activity in the low pH range and increasing activity with increasing pH values. Equally, the group of CES1 F and PPT1 displayed pH optima at strongly basic pH values, while here activity at lower pH was also relatively high.

[0222]

[0194] As the investigated hydrolases showed various pH-dependent activity profiles, this property might be harnessed to identify or exclude residual hydrolases in final drug products. To test this hypothesis, the ester cleaving activity of three model mAb formulations was analysed in the same way as in the above-described pH screenings (Figure 5B). Varying pH- dependent PS degrading activities were observed, which may indicate the presence of various residual hydrolases with unique pH optima. mAb formulation A showed strong activity towards higher (basic) pH values, but no enzymatic activity under acidic pH conditions. In contrast, mAb formulation B displayed strong similarities to the combined activity profiles of LPLA2 and LIPA. Lastly, mAb formulation C also showed strong activity towards higher pH values. However, unlike mAb A, mAb C revealed a medium activity in the acidic pH range, suggesting the presence of CES1 F and / or PPT1. In conclusion, differing pH activity profiles might be used to infer the presence or absence of specific hydrolases and could be a valuable resource when considering pH values of mAb formulation buffers.

[0223] Example 7: CHO hydrolase risk matrix

[0224]

[0195] In a final step, gathered data about actively PS-degrading hydrolases were summarized in a risk matrix. This can serve as a guideline for prioritizing potentially critical HCPs for targeted removal or monitoring during bioprocess development. Based on identified values for gene expression, occurrence in mAb products, cellular localization, and the PS degradation activity, risk points were assigned and ranked in a final score. All risk categories were weighted equally (Table 7). The pH dependent hydrolase activity was not incorporated into the risk matrix, as this factor is specific to the formulation buffer and has to be assessed individually. Of note, proteomic data was not part of the risk matrix, as quantitative proteomic datasets may not be readily available for biopharmaceutical production processes. In any case it was possible to demonstrate strong correlation between proteome and transcriptome datasets (supported by localization data). As a result of this assessment, top ranked and thus to be prioritized in removal during the bioprocess are the hydrolases LIPA, PPT1 , PAF-AH and LPLA2, which all showed strong PS degradation activity likely catalyzed by the Ser, His, Asp / Glu triad. Among these, LIPA was the only intracellular hydrolase and was also highly expressed on mRNA level thus rendering this enzyme particularly critical for bioprocesses suffering from low cell viability or high cell lysis at the end of the fermentation and / or during harvest. LPLA2 and PAF-AH both displayed high PS degradation capacity, for PAF-AH in combination with efficient secretion, which rendered it with the higher risk score. IAH1 has the lowest risk score as it combined medium PS activity, intracellular localization and comparably low mRNA expression. This risk matrix can be used as a guideline to focus on the most critical PS degrading HCPs in a CHO based manufacturing process and helps to ensure an efficient utilization of resources.

[0225] Table 7: Calculated values for risk matrix parameters: Expression, occurrence in monoclonal antibody (mAb) products, polysorbate (PS) degradation activity, localization. Rating approach is indicated in brackets in the top of the column. The mean of all values for a hydrolase resulted in the final score.

[0226]

[0196] 12 hydrolytic enzymes potentially acting on PS that were identified in 7 different mAb preparations and characterized. The results suggest that the presence of specific HCPs in purified mAbs is influenced by factors beyond gene expression levels, which is in line with literature on difficult-to-remove HCPs. Contaminating HCPs in final formulations persist after purification either via product association or co-elution due to product-resembling physicochemical protein properties.

[0227]

[0197] A CRISPR / Cas9 mediated LPL KO alone was not able to halt PS degradation completely.

[0198] It was successfully demonstrated that hydrolase containing crude cell culture supernatant can be applied in a PS degradation assay providing strong indications towards hydrolase specific PS degradation. In fact, only LIPA, IAH1 and CES1 did not confirm the results obtained with the purified hydrolases, which is probably due to their intracellular localization. The hydrolase SMPD1 was not active against PS. SMPD1 was successfully purified as confirmed by a size exclusion chromatogram and Western blot against the C- terminal His tag (data not shown). Furthermore, comparison of the separations of a SEC protein standard mixture and SMPD1 shows that the protein elutes according to the apparent molecular weight of the monomer (72 kDa), demonstrating that no aggregation has occurred (data not shown).

[0228]

[0199] Another important aspect is the localization of hydrolytic HCPs, especially in terms of CHO proteins as localization data are oftentimes not experimentally determined. Low harvest viabilities lead to the release of intracellular HCPs changing the HCP profile in the cell culture harvest including PS degrading enzymes. Thus, cell viability at harvest and cell-straining harvest procedures can be of particular importance in products suffering from predominantly intracellular HCPs.

[0229]

[0200] Enzymes active against PS were found to possess the classical catalytic triad, which consists of Serine (Ser), Histidine (His), and Aspartate / Glutamate (Asp / Glu). This finding is consistent with the activities of PS hydrolases investigated in previous research.9,13 However, acid ceramidase possessing the same catalytic triad does not exhibit PS degrading activity. 9,13 In contrast to the enzymes investigated here, acid ceramidase belongs to EC number 3.5 and not 3.1. Thus, the catalytic triad is a prerequisite for PS degradation but other protein properties are similarly important.

[0230]

[0201] Furthermore, we found that purified mAb formulations exhibited varying activity pH profiles, suggesting the involvement of multiple enzymes in PS degradation. Thus, the specific enzymes, which are active with regards to PS degradation, seem to depend on the downstream purification process and the specific monoclonal antibody product.

[0231]

[0202] The risk matrix allows to prioritize hydrolases for further investigation and potential mitigation strategies. LPL would most likely have scored the highest score in the risk matrix as it showed highest gene and protein expression values and was found in 6 out of 7 mAb products. In addition, LPL wt cell lines strongly degraded PS in comparison to LPL KO cell lines. Strikingly, LIPA, despite being mainly localized intracellularly, is frequently found in mAb preparations and scores the highest final score in the risk matrix. Approximately 10 % of LIPA is secreted or at least found in the cell culture supernatant. Considering its high expression level, it is conceivable that even at low secretion levels, relatively high concentrations of LIPA are reached in the cell culture supernatant comparable to or even higher than that of other PS hydrolases. Additionally, low harvest viabilities or increased cell lysis induced by the harvest method or necrosis might increase the release of intracellular LIPA, offering an easy-to-implement counter measurement. Although the risk matrix provides a guidance of which hydrolases should be more closely monitored in the final drug product, individual products can differ in their hydrolase profile. Therefore, an individual case assessment is necessary considering also other factors as the pH dependent activity profiles or the level of similarity towards the specific biopharmaceutical.

[0232] Due to the individual inherent features of the identified HCPs, multiple mitigation strategies are conceivable and must / can be combined to develop a PS-degradation-free final drug product. Stably HCP overexpressing CHO cells enabled the robust supply of HCPs, such as hydrolases, without a contamination by the strongly PS degrading LPL enzyme and facilitates the expansion of this dataset in the future.

[0233] Example 8: Generation and Analysis of 9x hydrolase KO

[0234]

[0203] The 9x hydrolase KO cells were generated using the knock-out strategy as described herein and as visualized in Figure 6. The resulting intermediate (Host LPL KO, LP KO, LPPL KO, LPPL KO, LPPLCC12 KO) and final clones (Multi KO, clones 1 D8, 1 F8, 1G1 , 2B4, 2C3 and 2G11) are summarized and described in Table 3 and the respective sgRNA used are listed in Table 4.

[0235]

[0204] The genomic excision of Carboxylesterase (Ces) cluster 1 (CES1 cluster) was performed in two sequential transfections yielding two excisions of 1007 kilobase pairs (kb) on one allele and a larger excision of 1629kb extending beyond the CES1 cluster on the other allele. The CES1 cluster is illustrated as a schematic arrangement of the different genes and their location in Figure 19A and in a simplified alignment of the genes within the CES1 cluster in Figure 19B. According to the best of our knowledge, this is the largest biallelic gene excision in CHO cells reported to date.

[0236]

[0205] As may be taken from Figure 19C / 20C only few Ces variants showed gene expression in CHO-K1 cells. However, this large-scale excision of the entire gene cluster removed all potential Ces variants, thereby avoiding putative functional reactivation of related Ces species.

[0237]

[0206] A similar approach was applied for Ces cluster 2 (506kb, CES2 cluster), where a single transfection yielded a biallelic Ces cluster 2 excision. The CES2 cluster is illustrated as a schematic arrangement of the different genes and their location in Figure 20A and in a simplified alignment of the genes within the CES2 cluster in Figure 20B. Also, in the CES2 cluster only a few Ces variants, primarily Ces2c (V4), showed gene expression in CHO-K1 cells as target PS-degrading hydrolase (Figure 20C). Thus, the excision of the CES1 cluster and the CES2 cluster together removed more than 1500 bp and can be even extended as shown by the even larger excision in the CES1 cluster on one allele.

[0238]

[0207] The sequential knockout (KO) of multiple hydrolases (Fig.6, Table 3) could be confirmed via gene expression analyses showing reduction or lack of transcription upon insertion and / or deletion (InDei) mutations or gene (cluster) excision for the intermediate clones (Fig. 7A) and 6 final 9x KO clones (Multi-hydrolase KO, Figure 7B). In addition, Ces cluster excision was confirmed by deletion PCRs using following primer pairs: For CES1 cluster: C1 C1 TGGAGCCTGGAGAGGAAAGA (SEQ ID NO: 43) + C1C2.1 GGTCAAGGGGCCTGATATCG (SEQ ID NO: 44), C1 C1 + C1C2.2

[0239] CCCATGCCACTGTGATCCTT (SEQ ID NO: 45) and for CES2 cluster: C2C1

[0240] GCAGGGAGGTGATTTCAGGG (SEQ ID NO: 46) + C2C2 AACAGCCACCGCTCTTAGTC (SEQ ID NO: 47). The 6 final 9x multi-hydrolase KO clones showed similarly reduced transcription levels of the target hydrolases (Figure 7B). A head-to-head fed-batch cultivation assessment of the parental CHO-K1 cell line with all intermediate (Figure 8A) and final hydrolase KO cell lines (Figure 8B) surprisingly showed acceptable and only slightly reduced peak VCDs and harvest viabilities.

[0241]

[0208] Subsequent FMA confirmed gradually reduced polysorbate (PS) degradation with increasing number of hydrolase KOs (Fig. 9) in HCCF. Particularly, knockout of the CES1 (LPPLCC2 KO) and the CES2 cluster (LPLCC12KO) considerably reduced PS degradation in HCCF, which was even further reduced when the genes coding for IAH1 and PPT1 knocked out in addition to obtain the final Multi-hydrolase KO clones.

[0242]

[0209] These results demonstrate that CHO cells with neglectable hydrolase activity due to a 9x hydrolase KO show only slightly reduced, but acceptable viability and VCD. This was unexpected and encouraging to use these engineered CHO cells for therapeutic protein production.

[0243]

[0210] Moreover, these cells comprise large genome excisions of > 1000 kb and > 500 kb comprising the entire CES1 cluster and the entire CES2 cluster. Thus, these results further show as a proof-of-concept that gene minimization strategies can be combined with cell engineering for protein production by not only inactivating or knocking-out singular genes, but by removing entire cluster or additional upstream and downstream regions to reduce the risk of functional reactivation of related genes within the excised locus, particularly where genes are arranged as cluster, and further to identify genomic locations that can be excised in the long term aim of substantial genome minimization in CHO cells in order to improve productivity. Excising large gene clusters reduces the genome content in the CHO cell and, thus, safes energy resources in terms of replication, transcription, translation and secretion.

[0211] Considering the surprising and encouraging results with the multiple-hydrolase KO clones, we next analyzed whether these cells are suitable for therapeutic protein productions, particularly with regard to productivity and protein quality.

[0244] Example 9: Generation and Analysis of Multi-hydrolase KO clones stably expressing lgG1 and lgG4 monoclonal antibodies

[0245]

[0212] A selection of final multi-KO cell lines was stably transfected with two representative monoclonal antibodies (lgG1 , lgG4) and resulting producer pools were cultivated in a fed- batch. Again, multi-KO producer pools showed slightly reduced peak VCDs and harvest viabilities in comparison to the parental CHO K1 producer pool for both, lgG1 (Figure 10A and 10B) and lgG4 (Fig. 11 A and 11 B). For most of the multi-KO producer pools volumetric productivity was slightly reduced (Figure 10C and 11C). Surprisingly, however, specific productivity was increased in comparison to the parental CHO K1 producer pool for lgG1 (Figure 10D) and lgG4 in the majority of clones (Fig.11 D). This was particularly encouraging, as this suggests that similar titers compared to the parental CHO K1 producer pool can be achieved with minor adaptations to the culture conditions and / or the medium or by improving viability by modifying susceptibility to apoptosis, e.g., by additionally knocking-out genes encoding pro-apoptotic proteins, such as BAX and / or BAK, or by introducing anti-apoptotic genes.

[0246]

[0213] Following fed-batch cultivation, mAbs were purified from the harvested cell culture fluid (HCCF) and HCCF and purified mAb fraction were analyzed. General measurement of host cell protein (HCP) content in HCCF and purified mAb fraction, showed, if at all, a slightly reduced HCP amount in the parental CHO K1 producer pools in comparison to the multi-KO producer pools (Fig.12).

[0247]

[0214] Hydrolytic activity in HCCF and purified mAb fraction was detected using the surrogate MLID4 assay and in HCCF PS reduction was measured using FMA assay at days 0, 1 , 3, 7 and 14. The MLID4 assay with HCCF and purified mAb fraction showed strongly reduced hydrolytic activity in multi-KO producer pools in HCCF (Fig.13A) and the purified mAb fraction (Figure 13B). These results were confirmed in a subsequent FMA assay in HCCF with strong PS degradation observed for parental CHO K1 producer pools and only neglectable degradation in multi-KO producer pools for lgG1 (Fig.14A) and lgG4 (Fig. 14B). FMA experiments for purified mAb fractions to confirm the results from the MLID4 assay are currently underway. In particular, the strong reduction of hydrolytic activity already in HCCF was promising, as this indicates that a great extent of PS degrading hydrolases is covered by the 9x multi KO. Example 10: Analysis quality attributes of lgG1 and lgG4 monoclonal antibodies produced in multiple-hydrolase KO clones

[0248]

[0215] Purity analyses of the purified mAb fraction showed no increase in high molecular weight species (HMW) for lgG1 (Figure 15A) and lgG4 (Figure 15B) produced in 6 different multi-hydrolase KO clones. Similarly, no increase in low molecular weight species (LMW) was observed for lgG1 (Figure 16A) and lgG4 (Figure 16B) produced in 6 different multihydrolase KO clones.

[0249]

[0216] Finally, charge heterogenicity of mAbs from multi-KO producer pools did not significantly alter for lgG1 (Figure 17A) and lgG4 (Figure 17B) in comparison to mAbs from the parental CHO K1 producer pool and also no difference in N-glycosylation of mAbs of the subtype lgG1 (Figure 18A) and lgG4 (Figure 18B) were observed. Of note, the similarity in product quality was highly unexpected as the 9x multi KO cell line underwent multiple rounds of transfection and single cell sorting, potentially influencing pathways responsible for product quality. Finally, this is an additional prove for the applicability of a multi-hydrolase KO cell line in future bioprocesses.

[0250] Example 11 : Additional Bax / Bakl knockout further improves bioprocess performance

[0251]

[0217] As demonstrated in Figures 10 and 11 , the final multi-KO clones all suffered from relatively low harvest viabilities at the end of a 14-day platform fed-batch process, which may lead to unintended leakage of HCPs into the harvested cell culture supernatant and thereby reducing productivity or impacting product quality (Obrstar et al., 2018, Analytical chemistry 90 (19), S. 11240-11247. DOI: 10.1021 / acs.analchem.8b01236; Tang et al., 2022 Tang et al., 2022, Biotechnology progress 38 (2), Artikel e3228, e3228. DOI: 10.1002 / btpr.3228). As low harvest viabilities are often caused by apoptotic cell death, we attempted to restore high harvest viabilities of the multi-KO host cell clones by apoptosis engineering via the additional KO of the pro-apoptotic signaling genes Bax and Bak1 (BB), which are coding for the apoptosis-inducing proteins “BCL2 associated X protein” and “Bcl-2 homologous antagonist / killer”, respectively. To date, multiple studies successfully applied BB KOs to generate apoptosis resistant cell lines with prolonged high viabilities and productivities (Cost et al., 2010, Biotechnology and bioengineering 105 (2), S. 330-340. DOI: 10.1002 / bit.22541 , Misaghi et al., 2013, Biotechnology progress 29 (3), S. 727-737. DOI: 10.1002 / btpr.1722; Tang et al., 2022). As a basis for the BB KO, multi-KO clone 2 G11 was selected since it showed overall highest product titers with comparable VCD and viability during an IgG production experiment (Figure 10 and 11). Similar to the hydrolase KOs, CRISPR / Cas9 was used to introduce InDeis in the respective Bax and Bak1 genes and single cell sorting was used to obtain clonal host cells. Successful KO was verified for Bak1 via qPCR demonstrating substantially reduced transcript levels in all 4 tested clones in comparison to the parental cell line(data not shown). As transcript levels of Bax were only reduced in one out of four clones, ICE analysis was performed and confirmed biallelic out-of-frame mutations for both targets (Bax and Bak1) in all four clones (data not shown). Additionally, successful KO was proved using a western blot analysis, in which occurrence of Bax and Bak1 protein bands were only detected in the parental CHO samples (data not shown).

[0252]

[0218] It was further tested whether harvest viabilities can be improved by an additional Bax / Bak1 knockout. The four final KO cell lines, including Bax / Bak1 (BB) and 9 critical hydrolase KOs, served as host cell line for the production of an IgG 1 and an lgG4 antibody format. Host cell lines were stably transfected using a transposon system and following GS- based selection a 14 days fed-batch was performed with the resulting mAb producing cell pools. In comparison to the parental producer pool, although peak VCDs for multi-KO producer pools were approx, halved with 10x106viable cells / ml, viabilities throughout the fed- batch were comparably high in all pools and harvest viabilities were >70% (Figure 21A / B). mAb titers were measured from d7 and showed a linear increase until day 14. Strikingly, final harvest titers of multi-KO producer pools were similar or even increased in comparison to the parental producer pools. Best performing producer pools (BB-multi-KO 1 and 4) yielded a final titer of >4500 mg / L and >8000 mg / L for lgG1 and lgG4, respectively (Figure 21C / D). Of note, increased volumetric productivity could be explained by drastically increased specific productivity (qp) values, which doubled for the multi-KO producer pools regardless of the expressed antibody format (Figure 21 E / F).

[0253]

[0219] As PS degradation potential in the harvested cell culture supernatant is an important indicator for hydrolase burden, which theoretically have to be later removed during downstream purification, an FMA assay in supernatant of day 14 harvest samples was performed. Similar to the results observed with supernatant of host cell cultivations, PS degradation was significantly decreased in all BB-multi-KO producer pool cultivations (Figure 22A / B), resulting in <10% PS degradation for BB-multi-KO samples in contrast to >50% PS degradation in supernatant derived from the parental CHO host cell line. This significantly reduced PS hydrolysis in all BB-multi-KO producer pools was confirmed regardless of antibody format and sample purity by additional 4-MUN assay measuring general hydrolytic activity in harvested cell culture supernatant (Figure 22C), and protein A purified mAb fractions (Figure 22D). While hydrolytic activity in crude supernatant could be reduced by >250-fold in samples derived from BB-multi-KO clones, purified mAbs still exhibited reduced hydrolytic activity by >45-fold as compared to material from the parental CHO cell line.

[0220] As repeated genome engineering over a long period of time combined with the high genomic plasticity of CHO cells harbors the risk of genome alterations influencing product quality attributes (Guo et al., 2023, Frontiers in bioengineering and biotechnology 11 , S. 1143157. DOI: 10.3389 / fbioe.2023.1143157; Wurm and Wurm, 2021 , Biotechnology journal 16 (7), e2100165. DOI: 10.1002 / biot.202100165), it was of utmost importance to evaluate this in the generated BB-multi-KO based producer pools. Product quality (PQ) characterizations were conducted after protein A purification and included the analysis of high molecular weight (HMW) species (Figure 23 A (lgG1) and B (lgG4)) and low molecular weight (LMW) species (Figure 23 C (IgG 1 ) and D (I gG4)), charge variants (Figure 23 E (I gG 1 ) and F (lgG4)) and N-glycosylation patterns (Figure 24). The vast majority of the product quality parameters were determined to be highly comparable to the material produced by the parental host cell line. Only a slight increase in mannosylation was detected in the material derived from the BB-multi-KO pools, resulting in a minimal shift of the overall N-glycan structure. Notably, high comparability was true for all tested BB-multi-KO derived producer pools demonstrating the applicability, validity and robustness of this multi-hydrolase plus Bax and Bak1 KO strategy. As the glycosylation pattern of stable pools and not clones were compared, single cell sorting will provide an additional opportunity to select producer clones with an optimal glycosylation phenotype. In summary, we generated a novel multi-hydrolase KO host cell line, which eventually represents a viable and competitive host cell line alternative with significantly reduced hydrolytic and thus PS degradation activity, while maintaining high product titers and product qualities for two tested mAb formats.

[0254]

[0221] To date, only a single study has employed a genomic KO strategy in mammalian cells to address PS degradation (Chiu et al., 2017). We could further expand this approach to the most relevant hydrolytic HCPs in CHO cells. Considering the dramatic expansion of complex formats (particularly antibody-derived molecules) in the drug development pipelines, PS degradation may become more challenging again as current process knowledge from standard IgG molecules and related downstream purification strategies might not be applicable. Novel formats possess different molecule characteristics and thus will require different downstream purification processes which may be more limited to only partial hydrolase removal and thus giving rise to increased PS degradation. Towards this end, our multi-hydrolase KO host cell line has the potential to enable the production of such complex molecules even at a higher productivity while assuring depletion of strongly PS-degrading lipases.

Claims

1. CLAIMS1. A CHO cell comprising, (i) a lipoprotein lipase (LPL) gene knockout, (ii) at least three further hydrolase gene knockouts, and additionally (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) of the CHO genome, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C).

2. The CHO cell of claim 1 , wherein at least one of the two or more genome excisions removes 1000 kb or more, preferably 1500 kb or more of the CHO genome.

3. The CHO cell of claim 1 or 2, wherein the LPL knockout, the at least three hydrolase knockouts and the two or more genome excisions are biallelic.

4. The CHO cell of any one of the preceding claims, wherein the two or more genome excisions are genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C).

5. The CHO cell of claim 4, wherein the two or more genome excisions are genome excisions (a) of the CES1 cluster removing at least 1000 kb and (b) of the CES2 cluster removing of at least 500 kb of the CHO genome.

6. The CHO cell of any one of the preceding claims, wherein the CHO cell further comprises a glutamine synthetase (GS) knockout and / or a BAX and BAK knockout.

7. The CHO cell of any one of the preceding claims, wherein gene knockout comprises gene excision and gene inactivation, preferably gene excision.

8. The CHO cell of any one of the preceding claims, wherein the (ii) at least three further hydrolase gene knockouts comprise at least three knockouts of genes coding for hydrolases (a) selected from the group consisting of isoamyl acetate-hydrolyzing esterase 1 (IAH1), lipase A (LIPA), palmitoyl-protein-thioesterase 1 (PPT1), platelet-activating factor acetylhydrolase (PAF-AH), and phospholipase A2 group XV (LPLA2); and / or (b) of the hydrolase class with the EC number 3.1 and comprising a catalytical triad consisting of a serine (Ser)-histidine (His)-aspartate / glutamate (Asp / Glu) motif.

9. The CHO cell of any one of the preceding claims, wherein the CHO cell further comprises a recombinant gene encoding a therapeutic protein.

10. A method of producing a recombinant protein, comprising,(a) introducing a nucleic acid comprising a gene coding for a therapeutic protein and a gene coding for a selection marker into the CHO cell of any one of claims 1 to 8;(b) culturing the cell of step (a) under conditions to produce the therapeutic protein;(c) harvesting the therapeutic protein; and(d) purifying and optionally formulating the therapeutic protein.

11. A pharmaceutical composition comprising a recombinant protein produced according to the method of claim 10.

12. A method of reducing polysorbate degrading activity in a CHO cell, comprising(a) providing a CHO cell;(b) knocking out (i) lipoprotein lipase (LPL) gene and (ii) at least three further hydrolase gene knockouts in the CHO cell; and(c) excising genomic regions comprising (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) in the CHO cell, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1 (CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C).

13. A method of reducing polysorbate degrading activity in a recombinant protein produced in a CHO cell, wherein the method comprises(a) providing a CHO cell;(b) knocking out (i) lipoprotein lipase (LPL) gene and (ii) at least three further hydrolase gene knockouts in the CHO cell;(c) excising genomic regions comprising (iii) two or more genome excisions deleting at least two gene clusters comprising at least one carboxylesterase gene, wherein each genome excision removes at least 500 kilobase pairs (kb) in the CHO cell, and wherein the two or more genome excisions comprise genome excisions of (a) a cluster (CES1 cluster) comprising the gene coding for carboxylesterase 1(CES1) and carboxylesterase 1f (CES1 F) and (b) a cluster (CES2 cluster) comprising the gene coding for carboxylic-ester hydrolase 2C (CES2C);(d) transfecting the CHO cell comprising the knockout of step (b) and the two or more genome excisions of step (c) with a gene coding for a recombinant protein, preferably a therapeutic protein, and a gene coding for a selection marker;(e) harvesting the therapeutic protein; and(f) purifying and optionally formulating the therapeutic protein.

14. The method of claim 12 or 13, wherein the CHO cell is the CHO cell according to claims 1-8.

15. Use of the cell of claim 9 for the production of a therapeutic protein.

Images