Artificial miRNA targeting multiple hydrolases

Abstract

The present invention relates to an artificial microRNA (artificial miRNA or amiRNA) comprising an artificial miRNA guide strand sequence that targets two or more mammalian hydrolases acting on ester linkages, where the two or more hydrolases are preferably endogenous hydrolases of Chinese Hamster Ovary (CHO) cells. The invention also relates to a mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA, and a mammalian cell (e.g., a CHO cell) comprising the artificial miRNA. The invention also relates to the use of said mammalian cell for producing a protein of interest, and to the use of artificial miRNA for preventing the expression of at least two endogenous hydrolases in a mammalian cell and / or the hydrolase activity associated with a recombinant protein produced in the mammalian cell. Furthermore, provided is a method for manufacturing a protein of interest comprising introducing said artificial miRNA into a mammalian cell and into a composition comprising a recombinant protein of interest and optionally a polysorbate wherein the recombinant protein of interest is obtained by said method.

Description

Technical Field

[0001] The present invention relates to an artificial microRNA (artificial miRNA or amiRNA), the artificial microRNA comprising an artificial miRNA guide strand sequence, the artificial miRNA guide strand sequence targeting two or more mammalian hydrolases acting on ester bonds, wherein the two or more hydrolases are preferably endogenous hydrolases of Chinese hamster ovary (CHO) cells. The present invention also relates to a mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA, and a mammalian cell (e.g., CHO cell) comprising the artificial miRNA. The present invention also relates to the use of the mammalian cell for producing a target protein, and the artificial miRNA for preventing the expression of at least two endogenous hydrolases in the mammalian cell and / or the hydrolase activity associated with the target recombinant protein produced in the mammalian cell. In addition, a method for manufacturing a target protein is provided, the method comprising introducing the artificial miRNA into a mammalian cell and into a composition comprising a target recombinant protein and optionally polysorbate, wherein the target recombinant protein is obtained by the method. Background Art

[0002] Many companies have observed polysorbate (PS) degradation and associated (sub)visible particle formation in biopharmaceutical formulations, which compromises drug product stability and ultimately poses a risk to the delivery of innovative medicines to patients. The main culprit for PS degradation is hydrolyzed host cell proteins (HCPs) originating from production cell lines, most of which are derived from Chinese hamster ovary (CHO) cells. Here, a small fraction of particularly difficult-to-remove HCPs (mainly lipases) leads to the hydrolytic cleavage of PS, resulting in the accumulation of free fatty acid aggregates / particles.

[0003] High product quality and stability are essential for biopharmaceutical drug products. Particle formation in formulated drug products (DP) such as monoclonal antibody (mAb) solutions has long been a concern in the biopharmaceutical industry, and enzymatic polysorbate (PS) degradation has been identified as a major underlying cause. Due to the structural similarity between PS (often used as a surfactant in DP formulations) and triglycerides (as natural substrates for hydrolases such as lipases acting on ester bonds), these enzymes are able to hydrolyze PS into polyoxyethylene (POE) and fatty acids, thus resulting in the formation of particles composed of fatty acids.

[0004] Particle formation in biopharmaceutical DP formulations has a negative impact on the long-term stability of the DP and therefore may pose a risk of causing an immune response in patients. In addition, some contaminating hydrolases / lipases (e.g., putative phospholipase B-like protein 2 (PLBL2)) represent non-human host cell proteins (HCPs) that are difficult to remove and have been reported to be immunogenic in patients and should therefore be removed or avoided in pharmaceutical DP compositions. HCPs that co-purify with the active pharmaceutical ingredient have been observed to vary among different biopharmaceuticals and, therefore, co-purifying contaminating hydrolases / lipases are believed to vary between biopharmaceutical products, cell lines, and manufacturing processes.

[0005] In recent years, several residual HCPs have been identified in pharmaceutical formulations and reported to be the likely cause of PS degradation. The identified enzymes mainly belong to the lipase class, a subclass of esterases that catalyze lipid hydrolysis (Yunyu et al., American Pharmaceutical Review, 2020). Among them, group XV lysosomal phospholipase A2 isoform X1 (LPLA2), putative phospholipase B-like protein 2 (PLBL2), liver carboxylesterase, group VII phospholipase A2 (PLA2G7), lysophospholipase 2 (LYPLA2), lysosomal acid lipase A and lipoprotein lipase (LPL) were detected in activity-based protein profiling assays and final drug products, indicating that these enzymes have potential key roles in PS degradation (Hall et al., Journal of Pharmaceutical Sciences, 2016, 105(5):1633-1642; Dixit et al., Journal of Pharmaceutical Sciences, 2016, 105(5):1657-1666; Zhang et al., Journal of Pharmaceutical Sciences, 2020, 109(11):3300-3307; Chiu et al., Biotechnology and Bioengineering, 2017, 114(5): 1006-1015; Li et al., Analytical Chemistry, 2021, 93(23): 8161-8169). However, not only lipases, but also other enzymes with PS degradation potential have been identified.

[0006] Li and colleagues reported the identification of 82 proteins using an activity-based protein spectrum method to enrich potential PS degrading enzymes. For example, sialyl acetylesterase (SIAE) was identified as a possible root cause of PS80 degradation in this study (Li et al., Analytical Chemistry, 2021, 93 (23): 8161-8169). Moreover, Graf et al., 2021 (Graf et al., Journal of Pharmaceutical Sciences, 2021, 110 (11): 3558-3567) describes palmitoyl protein thioesterase 1 (PPT1). Some HCPs that are difficult to remove show PS degradation characteristics, and are already present in trace amounts in formulated medicines. The presence of trace HCP contaminants in formulated drugs represents a major problem in terms of detectability of low-abundance HCPs and makes it more challenging to identify the lipase that actually causes PS hydrolysis (Hall et al., Journal of pharmaceutical sciences, 2016, 105 (5): 1633-1642; Zhang et al., Journal of Pharmaceutical Sciences, 2020, 109 (11): 3300-3307). In addition, Aboulaich et al. observed changes in HCP spectra associated with different types of mAb forms (Aboulaich et al., Biotechnology progress, 2014, 30 (5): 1114-1124). Similar results were reported for difficult-to-remove lipases, where lipases could be identified in mAb 1 preparations but not in mAb 2 preparations. This can be explained by the described binding behavior of lipases to the produced mAbs (Tran et al., Journal of chromatography A, 2016, 1438: 31-38; Li et al., Analytical Chemistry, 2021, 93 (23): 8161-8169). Therefore, individual lipases resist the purification process and may be responsible for PS degradation. In addition, the choice of the final production cell line also has an impact on the resulting HCP spectrum. Depending on the HCP expression profile of the selected production cell line, different amounts and compositions of HCPs bound by mAbs can be expected after the purification process and in the final drug product. In addition, batch-to-batch deviations regarding the final HCP content and composition are also common. Therefore, it is necessary to reduce the activity of contaminating lipases in biopharmaceuticals. Since contaminating lipases appear to vary from product to product, it is necessary to target multiple lipases simultaneously.Furthermore, some lipases (PLBL2) have been reported to be immunogenic in patients, so for this reason it would also be advantageous to remove these lipases, preferably before purification, ie already during cell line development and upstream processes.

[0007] One possible mitigation strategy is to remove such critical HCPs in production cell lines. The engineering of mammalian cell lines, and especially CHO cells, has a long history. Therefore, cell lines are altered by overexpression, downregulation or knockout (KO) of a single gene or a combination thereof to gain and / or lose a certain phenotype or characteristic, such as improving productivity, controlling product quality (such as glycosylation), and ensuring cell line stability. In addition to the more traditional knockout and knock-in methods, recent gene editing technologies have also become increasingly important. Another alternative is RNA interference (collectively referred to as small noncoding RNA), which leads to post-transcriptional control of mRNA.

[0008] Multigene regulation can be achieved via microRNA (miRNA). Various miRNA expression studies in CHO cells have shown that endogenous miRNAs are differentially expressed during different culture conditions (such as nutrient depletion), after temperature shifts, or at different culture stages (Druz et al., Biotechnology and Bioengineering, 2011, 108(7):1651-1661; Gammell et al., Journal of Biotechnology, 2007, 130(3):213-218; Hernández Bort et al., Biotechnology Journal, 2012, 7(4):500-515). In addition, by comparing the expression profiles of high- and low-producing CHO cells, natural miRNAs that affect recombinant protein expression and secretion were identified (Lin et al., Biotechnology Progress, 2011, 27(4): 1163-1171; Hammond et al., Biotechnology and Bioengineering, 2012, 109(6): 1371-1375; Maccani et al., Applied Microbiology and Biotechnology, 2014, 98(17): 7535-7548). In recent years, various human miRNAs have been shown to promote productivity and / or benefit cell growth in CHO cell lines. Therefore, a comprehensive miRNA mimic library was transiently transfected into CHO cells and the corresponding phenotypic changes were analyzed (Strotbek et al., Metabolic Engineering, 2013, 20: 157-166; Fischer et al., Biotechnol J. 2014 Oct; 9(10): 1279-1292). Therefore, miRNA candidates that have a positive impact on cell performance are stably co-expressed with different recombinant products.

[0009] During fed-batch culture, cell line engineering resulted in a product-independent increase in productivity, while no effect on product quality attributes was observed (Strotbek et al., Metabolic Engineering, 2013, 20: 157-166; Fischer et al., Biotechnology and Bioengineering, 2017, 114 (7): 1495-1510). These results not only demonstrate the feasibility of miRNA-based cell line engineering, but also demonstrate its high potential to generate genetically optimized CHO production cell lines. However, to date, CHO cell line engineering strategies using miRNAs have only been performed using naturally occurring miRNAs or their modifications (including cross-species applications).

[0010] In the early 1980s, the discovery of miRNA (Lee et al., Cell, 1993, 75 (5): 843-854) opened a new chapter in the field of gene regulation research, and it is speculated that half of the human transcriptome constitutes the basis of miRNA regulation. MiRNA guides the synergistic inhibition of multiple messenger RNA (mRNA) transcripts, thereby forming a complex gene regulatory network. Target recognition of miRNA is based on partial complementarity between miRNA and the 3' untranslated region (3'UTR) of each target mRNA, although targeting is also observed at a lower frequency in open reading frames (ORFs). Although miRNA exists naturally as a small RNA repressor of gene expression, artificial short interfering RNA (siRNA) has been developed as a powerful tool for experimental gene inhibition. In contrast to the partial complementarity of the observed miRNA-target interaction, siRNA is designed to be fully complementary to a single mRNA target, thereby being able to inhibit a single gene, and is well established and can be designed and generated as needed.

[0011] In order to realize the customized gene regulation of multiple specific target lipase, it is necessary to design autonomous design and non-natural artificial miRNA (amiRNA). The miRNA of such artificially designed target multiple expected mRNA is very rare, and has not yet been established. The de novo design of miRNA for target such as hydrolase (for example, lipase and / or carboxylesterase) is very difficult and has not yet been shown. In addition, although some natural miRNAs that improve such as productivity have been identified, it seems that there is no description or identification of targeting and the natural miRNA of a class of enzymes such as hydrolase that is lowered. Therefore, it is necessary to target and participate in the multiple hydrolases (comprising lipase and / or carboxylesterase) associated with immunogenicity simultaneously. Summary of the invention

[0012] Based on minimally conserved regions in the mRNA sequences of two sets of target HCPs, we provide proof-of-concept for simultaneous knockdown of multiple lipases in CHO cells using a single artificial miRNA (amiRNA). The designed amiRNA not only reduced PS degradation but also laid the foundation for expanding this tool to other areas of cell line phenotype engineering.

[0013] The object of the present invention is to provide a cell line engineering method for simultaneously targeting multiple hydrolases such as lipases acting on ester bonds in mammalian cells to reduce hydrolase activity, particularly PS degradation activity and / or immunogenic activity of contaminating hydrolases in a composition comprising a target recombinant protein (such as mAb) produced in the mammalian cells. Therefore, highly conserved transcripts of hydrolases (especially lipases) acting on ester bonds are targeted by artificial microRNAs (amiRNAs) designed from scratch. These amiRNAs simultaneously knock down multiple hydrolases such as lipases via complete and / or incomplete mRNA target binding, and can be used to stably or transiently express amiRNAs targeting multiple hydrolases (e.g., lipases) and target transgenes. The method of using the one or more amiRNAs to reduce hydrolase expression in mammalian cells, thereby resulting in reduced hydrolytic activity and thus reduced polysorbate degradation in products produced in mammalian cells, involves all unit operations of the manufacturing process, including the production process using cell culture (vial thawing until the harvest process, including HCCF), downstream intermediates and compositions containing the intermediates, and more final compositions (such as bulk drug substances (DS) and formulated drug products (DP)). In particular, the following proof of concept is provided: artificial miRNAs containing designed artificial miRNA guide strand sequences targeting multiple hydrolases (e.g., lipases or carboxylesterases) can be designed from scratch and effectively reduce and / or prevent PS degradation activity (due to the presence of contaminating HCPs with hydrolase activity acting on ester bonds) in harvested cell culture fluids of mammalian cells expressing a protein of interest. Since these artificial miRNAs can target multiple targets simultaneously and can be further used as a mixture of different artificial miRNAs, this method allows addressing PS degradation that has been contaminated in a composition containing a protein of interest in an upstream process without knowing the exact hydrolase that causes the observed degradation.

[0014] In one aspect, the present invention relates to an artificial microRNA (miRNA), the artificial microRNA comprising an artificial miRNA guide strand sequence, the artificial miRNA guide strand sequence targeting two or more mammalian hydrolases acting on ester bonds. The two or more hydrolases are endogenous hydrolases of mammalian cells, preferably human or rodent cells, more preferably CHO cells. The artificial miRNA guide strand sequence can have complete or incomplete complementarity with the target sequence of at least two mRNAs encoding two or more mammalian hydrolases. Preferably, the artificial miRNA guide strand sequence has incomplete complementarity with all target sequences of at least two mRNAs encoding two or more mammalian hydrolases or all target sequences except one target sequence. The target sequence is located in the 3' untranslated region (3'UTR) and / or the coding region of at least two mRNAs encoding two or more mammalian hydrolases, preferably in the coding region of at least two mRNAs encoding two or more mammalian hydrolases. In certain embodiments, the artificial miRNA targeting two or more mammalian hydrolases reduces the expression and / or activity of the two or more hydrolases in mammalian cells, preferably rodent or human cells, more preferably CHO cells.

[0015] The artificial miRNA guide strand sequence has about 19 to 25 nucleotides, preferably 20 to 24, more preferably 21 to 23 nucleotides. In addition, the artificial miRNA guide strand sequence independently hybridizes with the target sequences of at least two mRNAs encoding two or more mammalian hydrolases without forming a bulge loop and / or forms at most two bulge loops, preferably without forming a bulge loop. Preferably, the artificial miRNA guide strand sequence has at least 55% sequence complementarity with one or more target sequences of at least two mRNAs encoding two or more mammalian hydrolases; and / or wherein the miRNA guide strand sequence has at least 13 Watson-Crick base pairings with one or more target sequences of at least two mRNAs encoding two or more mammalian hydrolases. Examples of suitable miRNA guide strand sequences are artificial miRNA guide strand sequences having the nucleotide sequence of SEQ ID NO: 1 (LP 8mer), SEQ ID NO: 2 (LP 6mer), SEQ ID NO: 3 (LH 8mer), SEQ ID NO: 4 (LH 9mer) or SEQ ID NO: 5 (non-classical LH). Artificial miRNA can be pri-miRNA, pre-miRNA, double-stranded mature miRNA (including a follower strand and a guide strand) or a single-stranded miRNA guide strand. Artificial miRNA can also be a single-stranded or double-stranded miRNA mimic.

[0016] In certain embodiments, one of the two or more mammalian hydrolases is lipoprotein lipase (LPL). The one or more target sequences in the mRNA encoding LPL are preferably within nucleotides 560-590, 700-745, 930-960 or 1330-1360 of SEQ ID NO: 6 or 7, preferably within nucleotides 560-590, 705-745, or 930-960 of SEQ ID NO: 6 or 7. In additional or alternative embodiments, the two or more mammalian hydrolases are selected from lipoprotein lipase (LPL), phospholipase B-like protein 2 (PLBL2), phospholipase A1 (Pla1a), inactive pancreatic lipase-related protein 1 (PNLIPRP1), inactive pancreatic lipase-related protein 2 (PNLIPRP2), pancreatic triacylglycerol lipase precursor (PNLIP), lipase member I precursor (LipI), lipase member H precursor (LipH), lipase member G precursor (LipG), lipase member C precursor (LipC), and isoforms thereof.

[0017] In certain embodiments, the artificial miRNA guide strand sequence has incomplete complementarity with one or more target sequences of at least one of at least two mRNAs encoding two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) 8 nucleotides (nt 2-9), which is complementary to one or more target sequences and has 0, 1, 2 or 3 mismatches, preferably 0, 1 or 2 mismatches, and (ii) a GC content of 55-70%, wherein all or all except one or all except two guanines (G) or cytosines (C) are complementary to one or more target sequences; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides, which is complementary to one or more target sequences and has 1, 2 or 3 mismatches, preferably wherein the mismatches are not consecutive mismatches, and (ii) a GC content of ≤50% in the last 6 nucleotides, wherein 1 or 2 complementary nucleotides are G or C. Examples of such artificial miRNA guide strand sequences are artificial miRNA guide strand sequences having the nucleotide sequence of SEQ ID NO: 2 (LP 6mer), SEQ ID NO: 4 (LH 9mer) or SEQ ID NO: 5 (non-classical LH).

[0018] In other embodiments, the artificial miRNA guide strand sequence has incomplete complementarity with one or more target sequences of at least one of at least two mRNAs encoding two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) a sequence of 8 nucleotides (nt 2-9), wherein at least the first 4 consecutive nucleotides (at least nt 2-5) complementary to one or more target sequences, wherein the sequence of 8 nucleotides comprises at least one mismatch with the one or more target sequences, and (ii) a GC content of 55-75%, wherein at least 4 consecutive nucleotides complementary to the one or more target sequences are G or C; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides, which is complementary to the one or more target sequences and has 0 or 1 mismatch with the one or more target sequences, preferably wherein the last nucleotide is complementary to the one or more target sequences and / or the second to last nucleotide is a mismatch, and (ii) a GC content of 30-50% in the last 6 nucleotides, preferably wherein at least 2 complementary nucleotides are G or C. Examples of such artificial miRNA guide strand sequences are artificial miRNA guide strand sequences having a nucleotide sequence of SEQ ID NO: 1 (LP 8mer) or SEQ ID NO: 2 (LP 6mer).

[0019] In another aspect, the present invention provides a mammalian expression vector comprising a polynucleotide sequence encoding an artificial miRNA according to the present invention. The mammalian expression vector may further comprise at least one target gene encoding a target recombinant protein. In certain embodiments, the mammalian expression vector comprises a polynucleotide sequence encoding two or more artificial miRNAs according to the present invention, wherein the artificial miRNAs are the same or different.

[0020] In another aspect, the present invention relates to a mammalian cell comprising an artificial miRNA according to the present invention or a mammalian expression vector according to the present invention. The mammalian cell may be a human or rodent cell, preferably a rodent cell, more preferably a CHO cell. In a preferred embodiment, the mammalian cell also expresses a recombinant protein of interest.

[0021] In another aspect, the present invention relates to the use of an artificial miRNA according to the present invention in a mammalian cell for preventing the expression of at least two endogenous hydrolases in a mammalian cell and / or preventing the hydrolase activity and / or polysorbate degradation activity associated with a recombinant protein of interest produced in a mammalian cell.

[0022] On the other hand, a method for producing a target protein is provided, the method comprising the following steps: (a) introducing an expression vector comprising a target gene encoding a target recombinant protein into a mammalian cell; (b) introducing an artificial miRNA comprising an artificial miRNA guide strand sequence targeting two or more mammalian hydrolases acting on ester bonds into the mammalian cell, wherein the artificial miRNA is introduced into the mammalian cell before, simultaneously with or after step (a), wherein the artificial miRNA is introduced as RNA or as a mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA; (c) culturing the mammalian cell under conditions allowing expression of the target recombinant protein; and (d) harvesting the target recombinant protein in the harvested cell culture fluid. The method according to the present invention may also comprise the following steps: (e) purifying the target recombinant protein; and (f) formulating the purified target recombinant protein into a composition. The harvested cell culture fluid has reduced hydrolase activity and / or reduced PS degradation activity. Preferably, the harvested cell culture fluid has reduced hydrolase activity and / or reduced PS degradation activity compared to the harvested cell culture fluid containing the target recombinant protein produced in a mammalian cell that does not contain or encodes the artificial miRNA. In a preferred embodiment, the composition comprising the purified recombinant protein of interest has reduced hydrolase activity and / or reduced PS degradation activity.

[0023] In certain embodiments, the artificial miRNA is an artificial miRNA according to the present invention, and / or the mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA is a mammalian expression vector according to the present invention. The two or more mammalian hydrolases are endogenous hydrolases of mammalian cells, preferably the mammalian cells are human or rodent cells, more preferably CHO cells.

[0024] The present invention also encompasses a composition comprising a recombinant protein of interest, wherein the recombinant protein of interest can be obtained by the method of the present invention, and the composition preferably further comprises PS (such as PS20 or PS80), more preferably at a concentration of 0.1 g / L or higher. For example, in such a composition comprising PS (PS20 or PS80), preferably less than about 10% of the PS is degraded when the composition is stored at about 2°C to about 8°C for at least six months. The composition comprising the recombinant protein of interest has reduced hydrolase activity and / or reduced PS degradation activity, preferably compared to a composition comprising the recombinant protein of interest produced in mammalian cells without the introduction of or comprising artificial miRNA, the composition comprising the recombinant protein of interest has reduced hydrolase activity and / or reduced PS degradation activity. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1:Protein sequence alignment of LPL and its eight homologues (lipase 1a-1h) and LPL and PLBL2. The amino acid sequences of lipases from CHO host cell lines (LPL, lipase 1a-1h and PLBL2) were extracted from the database. The alignment was performed using Geneious The alignment type "Global alignment with free end gaps" of the software was used. (A) Alignment of amino acids 140-320 of LPL with its eight homologues (lipases 1a-1h). (B) Alignment of amino acids 240-450 of LPL with PLBL2. In common: black area = sequences conserved between all lipases; grey = sequences partially conserved between lipases; light grey = sequences less conserved between lipases. Identity: The long bars indicate the degree of homology between lipases. Lipase 1 = LPL; Lipase 2 = PLBL2; Lipase 1a = Pla1a; Lipase 1b = PNLIPRP1; Lipase 1c = PNLIPRP2; Lipase 1d = PNLIP; Lipase 1e = LipI; Lipase 1f = LipH; Lipase 1g = LipG; Lipase 1h = LipC.

[0026] Figure 2 : Transcript sequence alignment of LPL and its eight homologues (lipase 1a-1h) and LPL and PLBL2 for identification of amiRNA binding sites. Lipase transcript sequences (LPL, lipase 1a-1h and PLBL2) of CHO host cell lines were extracted from the database. The alignment was performed using Geneious The alignment type "global alignment with free end gaps" of the software was used. (A) Alignment of the transcript sequence 480-1020 of LPL with its eight homologs (lipases 1a-1h). Designed amiRNAs (grey arrows): 9 = LH 9mer; NC = non-classical LH; 8 = LH 8mer. (B) Alignment of the transcript sequence 620-1630 of LPL with PLBL2. Designed amiRNAs (grey arrows): 6 = LP 6mer; 8 = LP 8mer. In common: black area = sequence conserved between all lipases; grey = sequence partially conserved between lipases; light grey = less conserved sequence between lipases. Identity: The long bars indicate the degree of homology between lipases. Lipase 1 = LPL; Lipase 2 = PLBL2; Lipase 1a = Pla1a; Lipase 1b = PNLIPRP1; Lipase 1c = PNLIPRP2; Lipase 1d = PNLIP; Lipase 1e = LipI; Lipase 1f = LipH; Lipase 1g = LipG; Lipase 1h = LipC.

[0027] Figure 3 : De novo designed amiRNA targeting the binding of LPL and its eight homologues (lipase 1a-1h). Lipase transcript sequences (LPL, lipase 1a-1h) of CHO host cell lines were extracted from the database. Geneious The alignment type "global alignment with free end gaps" of the software was performed, resulting in the consensus sequence number indicated at the top of the figure. The designed amiRNAs are indicated by gray arrows. The goal of amiRNA design was 100% sequence complementarity to LPL (highest preference) and deviations are indicated. Black boxes indicate conserved nucleotides between all lipase targets and therefore 100% complementarity between amiRNA and its target, while mismatches between single and multiple lipases and amiRNAs are depicted as dots. (A) LH9mer, (B) LH NC, and (C) LH 8mer binding to LPL and its eight homologs. Lipase 1 = LPL; Lipase 2 = PLBL2; Lipase 1a = Pla1a; Lipase 1b = PNLIPRP1; Lipase 1c = PNLIPRP2; Lipase 1d = PNLIP; Lipase 1e = LipI; Lipase 1f = LipH; Lipase 1g = LipG; Lipase 1h = LipC.

[0028] Figure 4 : De novo designed amiRNA targeting the binding of LPL and PLBL2. Lipase transcript sequences of CHO host cell lines were extracted from the database (LPL = lipase 1, PLBL2 = lipase 2). Geneious The software's alignment type "global alignment with free end gaps" was performed to obtain the consensus sequence number indicated at the top of the figure. The designed amiRNA is indicated by a gray arrow. The amiRNA is designed to have 100% sequence complementarity to LPL (highest preference). The black box indicates the conserved nucleotides between all lipase targets and therefore 100% complementarity between the amiRNA and its target, while the mismatches between single and multiple lipases and amiRNAs are depicted with dots. Gaps / protrusions are indicated with dashes. Binding of LPL and PLBL2 to (A) LP 6mer, (B) LP 8mer.

[0029] Figure 5: Analysis of cell growth and viability of CHO host cells transiently transfected with amiRNA designed for various lipases. The CHO-K1 GS host cell line was transiently transfected with amiRNAs against LPL (lipase 1) and its eight homologues (LH) or LPL and PLBL2 (lipase 2) (LP). Cells were analyzed via flow cytometry 3 days after amiRNA transfection. (A) Viable cell concentration (VCC) and (B) viability of treated CHO host cells. NT siRNA = transfected with non-targeting siRNA, LP = amiRNA (6mer or 8mer) against LPL and PLBL2, LH = amiRNA (8mer, 9mer or non-classical) against LPL and its eight homologues, Mock = transfected with nuclease-free water, Tox5 = transfected with cell death-inducing siRNA. Lipase 1 (LPL) siRNA = LPL knockdown positive control. Biological replicates (n = 2) were measured. For statistical analysis, unpaired two-tailed t-test was applied (no*=not significant; *p≤0.05; ***p≤0.001).

[0030] Figure 6 : Normalized LPL concentration in harvested cell culture fluid (HCCF) samples of CHO host cells transiently transfected with amiRNA designed for multiple lipases. The CHO-K1 GS host cell line was transiently transfected with amiRNA for LPL (lipase 1) and its eight homologues (LH) or LPL and PLBL2 (lipase 2) (LP). HCCF samples were collected 3 days after transfection, and LPL concentration was measured by enzyme-linked immunosorbent assay. The measured LPL concentration was normalized (norm.) with viable cell concentration. NT siRNA = transfected with non-targeting siRNA, LP = amiRNA for LPL and PLBL2 (6mer or 8mer), LH = amiRNA for LPL and its eight homologues (8mer, 9mer or non-classical), mock = transfected with nuclease-free water, Tox5 = transfected with cell death-inducing siRNA, lipase 1 (LPL) siRNA = LPL knockdown positive control. Biological replicates (n = 2) were measured. For statistical analysis, unpaired two-tailed t-test was applied (no * = not significant; *p≤0.05).

[0031] Figure 7: Normalized PLBL2 concentration in HCCF samples of CHO host cells transiently transfected with amiRNA designed for various lipases. The CHO-K1 GS host cell line was transiently transfected with amiRNA for LPL (lipase 1) and its eight homologues (LH) or LPL and PLBL2 (lipase 2) (LP). HCCF samples were collected 3 days after transfection, and the PLBL2 concentration was measured by biofilm interferometry. The measured PLBL2 concentration was normalized with the viable cell concentration (norm.). NT siRNA = transfected with non-targeting siRNA, LP = amiRNA for LPL and PLBL2 (6mer or 8mer), LH = amiRNA for LPL and its eight homologues (8mer, 9mer or non-classical), mock = transfected with nuclease-free water, Tox5 = transfected with cell death-inducing siRNA. Biological replicates (n = 2) were measured. For statistical analysis, unpaired two-tailed t-test was applied (no*=not significant; *p≤0.05; **p≤0.01).

[0032] Figure 8 : Normalized Pla1a concentration in cell lysate samples of CHO host cells transiently transfected with amiRNA designed against various lipases. The CHO-K1 GS host cell line was transiently transfected with amiRNA against LPL (lipase 1) and its eight homologues (LH) or LPL and PLBL2 (lipase 2) (LP). Cell lysate samples were prepared 3 days after transfection, and Pla1a (lipase 1a) concentration was measured by enzyme-linked immunosorbent assay. The measured lipase 1a concentration was normalized (norm.) with the total protein concentration determined by Bradford assay. NT siRNA = transfected with non-targeting siRNA. LP = amiRNA against LPL and PLBL2 (6mer or 8mer), LH = amiRNA against LPL and its eight homologues (8mer, 9mer or non-classical), Mock = transfected with nuclease-free water, Tox5 = transfected with cell death-inducing siRNA. Biological replicates (n = 2) were measured. For statistical analysis, unpaired two-tailed t-test was applied (no * = not significant; *p≤0.05).

[0033] Fig. 9 : Root cause analysis of nonspecific lipase target knockdown mediated by amiRNA. Lipase transcript sequences of CHO host cell lines (Pla1a = lipase 1a, PLBL2 = lipase 2) were extracted from the database. Geneious The "Add Primers to Sequence" function of the software was used to identify potential amiRNA binding sites. (A) Off-target analysis of LP 8mer amiRNA in lipase 1a transcript sequence. (B) Off-target analysis of LH 9mer amiRNA in PLBL2 transcript sequence. AmiRNAs are indicated by gray arrows. Black boxes indicate nucleotides that are complementary to mRNA, while mismatches between lipase transcript sequence and amiRNA are depicted with dots.

[0034] Figure 10: Analysis of polysorbate degradation in HCCF samples of CHO host cells transiently transfected with amiRNA designed for various lipases. CHO host cells were transiently transfected with de novo amiRNA for post-translational regulation of LPL (lipase 1) and PLBL2 (lipase 2) (LP) or LPL and its eight homologs (LH). HCCF samples were collected 3 days after transfection. PS20 or PS80 was spiked into HCCF samples and incubated in the dark at room temperature for 0, 1, 3, 7 and 14 days. PS concentration was measured by fluorescent micelle assay (FMA). (A) PS20 concentration of HCCF samples (initial PS20 concentration 0.4 mg / ml). (B) PS80 concentration of HCCF samples (initial PS80 concentration 0.2 mg / ml). NT siRNA = transfection with non-targeting siRNA, LP = amiRNA (6mer or 8mer) against LPL and PLBL2, LH = amiRNA (8mer, 9mer or non-classical) against LPL and its eight homologues, Mock = transfection with nuclease-free water, Lipase 1 (LPL) siRNA = LPL knockdown positive control. Biological replicates (n = 2) were measured. For statistical analysis, unpaired two-tailed t-tests were applied (none * = not significant; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001).

[0035] Fig.11: Average polysorbate degradation rates in HCCF samples of CHO host cells transiently transfected with amiRNA designed against various lipases. Average PS20 and PS80 degradation rates were calculated based on FMA data generated from HCCF samples of CHO host cells transfected with amiRNAs. The amiRNAs targeted LPL (lipase 1) and PLBL2 (lipase 2) (LP) or LPL and its eight homologues (LH). The initial PS concentration was subtracted from the final measured PS concentration and divided by the incubation time to calculate the degradation rate per day. NT siRNA = transfected with non-targeting siRNA, LP = amiRNA (6mer or 8mer) against LPL and PLBL2, LH = amiRNA (8mer, 9mer or non-classical) against LPL and its eight homologues, mock = transfected with nuclease-free water, lipase 1 (LPL) siRNA = LPL knockdown positive control. Biological replicates (n = 2) were measured. For statistical analysis, unpaired two-tailed t-test was applied (no* = not significant; *p≤0.05; **p≤0.01; ***p≤0.001).

[0036] Figure 12: Analysis of viability and growth performance of IgG1 or IgG4 CHO production cells transiently transfected with amiRNA designed against various lipases. IgG1 (top) or IgG4 (bottom) CHO production cell lines were transiently transfected with amiRNA against LPL (lipase 1) and PLA1A (lipase 1a) (LH) or LPL and PLBL2 (lipase 2) (LP). Cells were cultured in batches and analyzed via flow cytometry at 3, 4 and 5 days after transfection. (A) Viability 5 days after transfection. (B) Integrated viable cell concentration (IVC) on day 5. NT siRNA = transfected with non-targeting siRNA, LP = amiRNA (6mer or 8mer) against LPL and PLBL2, LH = amiRNA (9mer, 8mer or non-classical) against LPL and PLA1A, Mock = transfected with nuclease-free water, Tox5 = transfected with cell death-inducing siRNA. Biological replicates (n=2) were measured. For statistical analysis, an unpaired two-tailed t-test 627 was applied (none* = not significant; *p≤0.05; **p≤0.01; ***p≤0.001).

[0037] Figure 13: Analysis of cell specific productivity and product concentration of IgG1 or IgG4 CHO production cells transiently transfected with amiRNA designed against various lipases. IgG1 (top) or IgG4 (bottom) CHO production cells were transiently transfected with amiRNA against LPL (lipase 1) and PLA1A (lipase 1a) (LH) or LPL and PLBL2 (lipase 2) (LP). HCCF samples were collected 5 days after transfection and product concentrations were determined using protein A-based biofilm interferometry. Specific productivity (QP) was calculated using IVC and titer data (in pg / cell*day (pg / c*d)). (A) Specific (spec.) productivity (pg / cell / day) 5 days after transfection. (B) Product concentration (mg / l) 5 days after transfection. NT siRNA = transfected with non-targeting siRNA, LP = amiRNA (6mer or 8mer) against LPL and PLBL2, LH = amiRNA (9mer, 8mer or non-canonical) against LPL and PLA1A, Mock = transfected with nuclease-free water, Tox5 = transfected with cell death-inducing siRNA. Biological replicates (n=2) were measured. DETAILED DESCRIPTION

[0038] 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 feature, element, integer, step, or component, but does not exclude the presence or addition of one or more other features, elements, integers, steps, components, or groups thereof. The term "comprising" therefore includes the more restrictive terms "consisting of" and "consisting essentially of". With respect to sequences, the terms "a nucleotide / amino acid sequence having" and "a nucleotide / amino acid sequence comprising" are used interchangeably, and include the embodiment "consisting of a 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. In addition, the singular and plural forms are not used in a restrictive manner. As used herein, the singular forms "a", "a kind of", and "the" specify both the singular and the plural, unless explicitly stated to specify only the singular.

[0039] The term "protein" is used interchangeably with "amino acid sequence" or "polypeptide" and refers to a polymer of amino acids of any length. These terms also include proteins that are post-translationally modified by reactions including, but not limited to, glycosylation, acetylation, phosphorylation, glycation, or protein processing. Modifications and changes can be made in the structure of a polypeptide, for example, fusion with other proteins, amino acid sequence substitutions, deletions, or insertions, while the molecule maintains its biological functional activity. For example, specific amino acid sequence substitutions can be made in a polypeptide or its underlying nucleic acid coding sequence, and proteins with the same properties can be obtained. Proteins produced or manufactured in mammalian cells or encoded by mammalian expression vectors (particularly using artificial miRNAs according to the present invention) are also referred to as target proteins. Preferably, the target protein is a target recombinant protein.

[0040] As used herein, the term "recombinant protein" refers to a protein generated by recombinant techniques such as molecular cloning, and may also be referred to as a target recombinant protein. As used herein, a recombinant protein is, for example, a target protein to be purified in a sample or to be produced in a mammalian host cell. Recombinant techniques bring together genetic material from multiple sources or create non-natural sequences. Recombinant proteins are typically based on sequences from cells or organisms or different species different from the receptor host cells (e.g., CHO cells or human embryonic kidney (HEK) 293 cells) used to produce proteins in cell culture, or based on artificial sequences, such as fusion proteins. In the context of the present invention, target recombinant proteins are preferably therapeutic proteins, such as antibodies, antibody fragments, antibody-derived molecules (e.g., scFv, dual or multi-specific antibodies) or fusion proteins (e.g., Fc fusion proteins). Therefore, in one embodiment, target recombinant proteins are selected from antibodies, antibody fragments, antibody-derived molecules, glycoproteins and fusion proteins.

[0041] 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 expression vectors (particularly plasmids) and integration into the genome of a host cell, it will be understood by those skilled in the art that it refers to a DNA sequence or molecule. In the context of miRNA, it will be understood by those skilled in the art that it refers to an RNA sequence or molecule or a DNA sequence encoding a miRNA.

[0042] As used herein, the term "mammalian cell" refers to a cell derived from a mammal, in particular a cell line. In the present invention, "mammalian cell" particularly encompasses human or rodent cells, such as HEK 293 cells or derivatives thereof or CHO cells and derivatives thereof. In most cases, it refers to CHO cells or derivatives thereof. Cells as referred to herein are cells maintained in culture and do not relate to primary cells, but rather to cell lines or cell line-derived cells, i.e., immortalized cells.

[0043] As used herein, the term "drug substance" (abbreviated as DS) refers to an active pharmaceutical ingredient (API) formulated with excipients. In contrast to excipients that aid in the delivery of the API, the API has a therapeutic effect in vivo. In the case of biotherapeutics, 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 called drug product).

[0044] As used herein, the term "drug product" (abbreviated as DP) refers to the final marketed dosage form of the drug substance, such as a tablet or capsule, or in the case of a biologic, typically a solution for injection in an appropriate container (such as a vial or syringe). The drug product may also be in lyophilized form. The antibody is preferably provided as an aqueous formulation in a glass vial or glass syringe.

[0045] As used herein, the term "polysorbate 20" refers to the following nonionic polysorbate-type surfactants: lauric acid esters of sorbitol and its anhydride copolymerized with approximately 20 moles of ethylene oxide per mole of sorbitol and sorbitan (polyoxyethylene (20) sorbitan monolaurate; CAS No.: 9005-64-5), such as The United States Pharmacopeial Convention (The United States Pharmacopeial Convention) (officially effective on May 1, 2020) is defined. It is also called Tween20. Its stability and relative non-toxicity allow it to be used as a surfactant and emulsifier in many domestic scientific analyses. Polysorbate 20 can be used as a detergent in immunoassays, Western blotting and enzyme-linked immunosorbent assays (ELISA). It can be further used in pharmacological applications such as pharmaceutical preparations, particularly for biological preparations, such as antibodies and Fc fusion proteins. In particular, it helps to prevent nonspecific antibody binding interactions.

[0046] As used herein, the term "polysorbate 80" refers to a nonionic polysorbate-type surfactant that is a mixture of partial esters of fatty acids (primarily oleic acid) with sorbitol and its anhydrides, which are ethoxylated with approximately 20 moles of ethylene oxide per mole of sorbitol and sorbitan anhydride (polyoxyethylene (20) sorbitan monooleate, CAS No.: 9005-65-6), such as It is defined in the United States Pharmacopeia Convention (Phase 6 Unified Edition, officially effective May 1, 2016). It is also called Tween 80 and has similar uses to polysorbate 20.

[0047] As used herein, the term "contaminating" or "contamination" refers to the presence of undesirable and / or unintentional substances, such as substances that are immunogenic and / or have PS-degrading enzyme activity, for example, enzymes HCPs with lipolytic activity and / or at least one protein or substance with hydrolytic activity (such as lipase activity), which can only be regarded as trace components compared to other proteins mainly produced for medical treatment, such as antibodies or antibody-like compounds. In the context of the present invention, hydrolytic activity and in particular lipase activity are undesirable due to their PS-degrading potential and may be co-purified with the recombinant protein. This applies in particular to the final formulated protein preparation, which advantageously contains only less than 1% (w / w), preferably less than 0.1% (w / w), more preferably less than 0.01% (w / w) of such undesirable factors compared to the total protein content (i.e., in the drug product).

[0048] As used herein, the term "polysorbate degrading enzyme activity" or "polysorbate degrading activity" refers to the activity of a substance (typically a host cell derived protein (enzyme)) that catalyzes the hydrolysis of ester bonds in PS. This includes HPC, particularly enzymes with hydrolytic activity such as lipase activity and / or carboxylesterase activity. As used herein, the term "lipase activity" refers to the activity of a protein (enzyme) that catalyzes the hydrolysis of ester bonds in lipids (such as fatty acid esters). As used herein, the term "carboxylesterase activity" refers to the activity of a substance (typically a protein (enzyme)) that catalyzes the hydrolysis of ester bonds in carboxylates. Lipase or carboxylesterase is a hydrolase that cleaves esters into acids and alcohols in a chemical reaction (also referred to as hydrolysis) with water. Many lipases and carboxylesterases belong to the carboxylester hydrolase category (EC 3.1.1). Although carboxylesterases form a separate carboxylesterase class (EC 3.1.1.1), the lipase may be, but is not limited to, a triacylglycerol lipase (EC 3.1.1.3), a phospholipase A2 (EC 3.1.1.4), a lysophospholipase (EC 3.1.1.5), (EC 3.1.1.23), a galactolipase (EC 3.1.1.26), a phospholipase A1 (EC 3.1.1.32), a lipoprotein lipase (EC 3.1.1.34) or a hormone-sensitive lipase (EC 3.1.1.79); a phosphodiesterase (EC 3.1.4) such as a phospholipase D (EC 3.1.4.4), a phosphatidylinositol phospholipase C (EC 3.1.1.5). 3.1.4.11), glycosylphosphatidylinositol phospholipase D (EC 3.1.4.50) or N-acetylphosphatidylethanolamine-hydrolyzing phospholipase D (EC 3.1.4.54); or glycosphingolipid deacylase (EC 3.5.1.69). As used herein, the term "hydrolase activity" is a more general term, including lipases, but also refers to the hydrolysis of compounds other than lipids, such as sialyl acetylesterase (SIAE) that catalyzes the removal of O-acetyl ester groups from sialic acid by hydrolysis, and thioester hydrolases (EC 3.1.2) such as palmitoyl protein thioesterase 1 (PPT1). As used herein, the term "hydrolase acting on ester bonds" relates to enzymes (particularly lipases or carboxylesterases) having hydrolase activity that catalyzes the hydrolysis of ester bonds, wherein "acting on ester bonds" means cleaving ester bonds.

[0049] As used herein, the term "therapeutic protein" refers to a protein that can be used for medical treatment of humans and / or animals. These include, but are not limited to, antibodies, growth factors, coagulation factors, vaccines, interferons, hormones, and fusion proteins. In the context of the present invention, a therapeutic protein is an antibody, including antibody-derived molecules.

[0050] As used herein, the term "produced" refers to the production of a target recombinant protein, particularly an antibody, in a mammalian cell (particularly a CHO cell) in a cell culture. Those skilled in the art know how to use fermentation to produce recombinant proteins in cells, and particularly antibodies. The production of recombinant proteins includes culturing mammalian cells, such as CHO cells, expressing the target recombinant protein in a cell culture. Cultivating cells expressing the recombinant protein in a cell culture includes maintaining the cells in a suitable culture medium and under conditions that allow growth and / or protein production / expression. The target recombinant protein can be produced by fed-batch or continuous cell culture. Therefore, mammalian cells such as CHO cells can be cultured in fed-batch or continuous cell culture or a combination thereof, preferably in fed-batch cell culture.

[0051] The term "expressing a recombinant protein of interest" or "expressing an antibody" refers to a cell that contains a DNA sequence encoding a recombinant protein of interest (such as an antibody), which DNA sequence is transcribed and translated into a 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.

[0052] As used herein, the term "about" refers to a 10% variation of the specified value, for example, about 50% includes variations from 45% to 55%.

[0053] The terms "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 other equivalents of numbers, such as three or more or four or more.

[0054] As used herein, the term "yield" refers to the amount of antibody after purification relative to the amount of antibody before purification (such as in the starting material). The term "step yield" refers to the amount of antibody after a purification step relative to the amount of antibody before the purification step.

[0055] As used herein, the term "siRNA" refers to a short double-stranded RNA typically consisting of 21-23 base pairs that are completely complementary to its target mRNA. siRNA usually contains two nucleotide overhangs at the 3' end, but a variety of changes in length and overhangs are allowed. siRNA consists of an active (guide) strand and a complementary inactive (follower) strand. siRNA is considered to be an exogenous RNA that enters the endogenous RNAi pathway. In contrast to endogenous miRNA, exogenous siRNA has no nuclear phase during processing, but enters the same processing pathway as dsRNA in the Dicer association step. As an alternative, siRNA can be introduced into cells by using a vector to express small hairpin RNA (shRNA) (also called short hairpin RNA). shRNA is a precursor of siRNA. shRNA enters the endogenous RNAi machinery to be processed into single-stranded siRNA.

[0056] As used herein, the term "miRNA" or "microRNA" refers to endogenous RNA produced from cells, and has partial or incomplete complementarity with their target mRNA, and binds to multiple mRNAs to inhibit their expression. MiRNA is a naturally occurring RNA that is transcribed from their genes by RNA polymerase II. MiRNA is naturally expressed as a long primary miRNA transcript (pri-miR) by miRNA genes. After being partially cut by the microprocessor complex in the nucleus, the stem-loop precursor miRNA (pre-miRNA) is exported to the cytoplasm, where it is further processed by Dicer into a double-stranded RNA consisting of an active or mature chain and an inactive follower chain. The loop region of the precursor miRNA is removed, and a 21-24bp double-stranded RNA (dsRNA) consisting of a follower chain and a guide chain is produced. The follower chain has no additional function and is therefore degraded by the cell. Mature miRNA (miRNA guide chain) is introduced into the RNA-induced silencing complex (RISC) to initiate gene silencing. In contrast to the perfect complementarity between siRNA and its target mRNA, miRNA has incomplete complementarity, binding primarily through the seed region sequence at its 5′ end. The key difference between siRNA and miRNA is that siRNA is specific for a single target site in a single mRNA, thus inhibiting the expression of one target gene, whereas miRNA has multiple targets and can regulate multiple genes, since mRNA recognition requires binding to a much shorter seed region sequence rather than the entire 21-23 nucleotide sequence of the siRNA. To initiate RNAi, siRNA must be fully complementary to its target mRNA, whereas miRNA can be partially complementary and thus bind to multiple mRNAs to inhibit their expression, and their mechanisms of action are different: siRNA cleaves mRNA, whereas miRNA inhibits the translation of mRNA and / or initiates the degradation of mRNA.

[0057] As used herein, the term "artificial microRNA", "artificial miRNA" or "amiRNA" refers to an artificial RNA comprising a de novo designed (i.e., computer designed, non-natural) guide strand sequence that binds to a variety of mRNAs to inhibit their expression. As used herein, an artificial miRNA binds to at least one natural mammalian cell mRNA with incomplete complementarity to inhibit its expression. Artificial miRNAs can be designed as double-stranded RNA (dsRNA) or mature miRNA (miRNA guide strand) by using a natural or artificial scaffold. Artificial miRNAs can also be transcribed by RNA polymerase III. Although the term artificial miRNA is sometimes also used for recombinant or synthetically produced miRNAs comprising natural miRNA guide strand sequences, the artificial miRNA according to the present invention comprises an artificial miRNA guide strand sequence, i.e., a de novo designed, non-naturally occurring (previously unidentified) miRNA guide strand sequence.

[0058] As used herein, the term "expression" refers to the transcription and / or translation of heterologous polynucleotide sequences in host cells. The expression level of a polynucleotide sequence (such as a polynucleotide sequence encoding a hydrolase or a target protein acting on an ester bond in a host cell) can be determined based on the amount of the corresponding mRNA present in the cell or by the amount of the desired polypeptide / target protein encoded by the selected sequence as in the embodiment of the present invention. For example, the mRNA transcribed by the selected sequence can be quantitatively measured by Northern blot hybridization with cell RNA, ribonuclease RNA protection, in situ hybridization or by polymerase chain reaction (PCR). The protein encoded by the selected sequence can be quantitatively measured by various methods, for example, by ELISA, by Western blot, by radioimmunoassay, by immunoprecipitation, by measuring the biological activity of the protein, by immunostaining of the protein and then performing flow cytometry analysis or by homogeneous time-resolved fluorescence (HTRF) measurement. The expression level of non-coding RNA (such as miRNA) can be quantitatively measured by PCR (such as qPCR).

[0059] As used herein, the term "harvested cell culture fluid" refers to the cell culture fluid (CCF) from which cells are removed during harvest, typically by centrifugation or filtration. Thus, particularly for secreted proteins, harvested cell culture fluid refers to the cell culture supernatant after harvest (i.e., cell removal).

[0060] In this proof-of-concept study, three different model lipases, including lipoprotein lipase (LPL), putative phospholipase B-like protein 2 (PLBL2), and phospholipase A1 member A (PLA1A), were selected based on their PS degradation potential reported in the literature and their presence in internal pharmaceutical mass spectrometry analysis (Chiu et al., 2017, Biotechnology and Bioengineering, 114(5), 1006-1015; Dixit et al., 2016, Journal of Pharmaceutical Sciences, 105(5), 1657-1666; Graf et al., 2021, Journal of Pharmaceutical Sciences, 110(11), 3558-3567; Li et al., 2021, Analytical Chemistry, 93(23), 8161-8169). The application of computer-designed, non-naturally occurring amiRNAs to simultaneously knock down multiple lipases in CHO cells, thereby reducing PS degradation, is described. Therefore, micro-conserved regions among the mRNA sequences of LPL, PLA1A and PLBL2 were exploited to identify target sites for simultaneous amiRNA binding.Novel amiRNA technology is a novel and promising approach in cell line engineering for targeted pathway engineering based on binding to conserved sequences of target transcripts.

[0061] Artificial microRNAs (miRNAs) targeting two or more mammalian hydrolases

[0062] In one aspect, the present invention relates to an artificial microRNA (miRNA), the artificial microRNA comprising an artificial miRNA guide strand sequence, the artificial miRNA guide strand sequence targeting two or more mammalian hydrolases acting on ester bonds. Targeting two or more mammalian hydrolases acting on ester bonds means respectively binding to two or more mRNAs encoding two or more mammalian hydrolases acting on ester bonds, and inhibiting their expression in mammalian cells (host cells). In other words, the artificial miRNA guide strand binds to two or more mRNAs encoding mammalian hydrolases acting on ester bonds, and inhibits their expression in mammalian cells (host cells). Two or more hydrolases are endogenous hydrolases of mammalian cells. Preferably, mammalian cells are rodent cells or human cells, and more preferably mammalian cells are rodent cells, such as CHO cells. Therefore, in certain embodiments, two or more mammalian hydrolases acting on ester bonds are rodent or human hydrolases, and preferably two or more mammalian hydrolases are rodent hydrolases such as hydrolases from CHO cells. In certain embodiments, the two or more mammalian hydrolases are three or more mammalian hydrolases, four or more mammalian hydrolases, five or more mammalian hydrolases, six or more mammalian hydrolases, seven or more mammalian hydrolases, or eight or more mammalian hydrolases.

[0063] The artificial miRNA guide strand sequence has complete or incomplete complementarity with the target sequence of at least two mRNAs encoding two or more mammalian hydrolases. Two or more mammalian hydrolases may also be referred to as two or more targeted mammalian hydrolases. The artificial miRNA guide strand sequence has incomplete complementarity with at least one of the at least two mRNAs encoding two or more mammalian hydrolases, preferably more than one of the target sequences of at least two mRNAs encoding two or more mammalian hydrolases. In one embodiment, the artificial miRNA guide strand sequence has complete complementarity with the target sequence of one mRNA in at least two mRNAs encoding two or more mammalian hydrolases. Therefore, the artificial miRNA guide strand sequence has complete complementarity with the target sequence of one mRNA of at least two mRNAs encoding two or more mammalian hydrolases, and has incomplete complementarity with one or more target sequences of the remaining mRNAs of at least two mRNAs encoding two or more (targeted) mammalian hydrolases. In certain embodiments, the artificial miRNA guide strand has incomplete complementarity with one or more target sequences of at least two mRNAs encoding two or more (targeted) mammalian hydrolases, or the artificial miRNA guide strand has incomplete complementarity with one or more target sequences of at least two mRNAs encoding two or more (targeted) mammalian hydrolases. Preferably, the artificial microRNA (miRNA) according to the present invention comprises an artificial miRNA guide strand sequence targeting three, four, five, six, seven, eight, nine or more mammalian hydrolases acting on ester bonds. In certain embodiments, the artificial miRNA guide strand has incomplete complementarity with one or more target sequences of all or all mRNAs encoding three, four, five, six, seven, eight, nine or more (targeted) mammalian hydrolases except one. It should be understood by those skilled in the art that the artificial miRNA guide strand sequence can have complete complementarity with the target sequence of one mRNA in the mRNA encoding three, four, five, six, seven, eight, nine or more mammalian hydrolases.

[0064] As used herein, the term "complete complementarity" means that in two sequences, when they are aligned and / or hybridized in antiparallel to each other, the nucleotide bases at each position in the sequence are complementary (no mismatches). It may also be referred to as "complete complementarity" or "complete complementarity". In the context of the present invention, the term "complementarity" is used for the antiparallel alignment of RNA strand sequences, especially between miRNA guide strand sequences and mRNA target sequences. Given that there are four choices for each base in the strand, the 20bp-22bp length of the guide strand sequence results in more than 4×10 12possible combinations. Considering that the length of the human genome is about 3.1 billion bases (3.1×10 12 bases), which means that each 20 bp to 22 bp miRNA guide strand sequence should only find one fully complementary match in the entire human genome.

[0065] The terms "incomplete complementarity" or "partial complementarity" are used synonymously herein and mean that in two sequences, when they are aligned and / or hybridized in antiparallel relation to each other, one or more nucleotide bases at at least one position in the sequence will not be complementary (one or more mismatches). As used herein, the term "alignment" refers to a theoretical way of arranging sequences, while as used herein, the term "hybridization" refers to the phenomenon of annealing of a single-stranded DNA or RNA molecule to a complementary DNA or RNA. In other words, incomplete complementarity describes two sequences that hybridize under physiological conditions and contain one or more mismatches. Incomplete complementarity can also be defined by less than 100% sequence complementarity, such as 96% or less sequence complementarity, such as 50% to 96%, preferably 55% to 96% sequence complementarity, and more preferably 60 to 96% sequence complementarity.

[0066] The target sequence may be located anywhere in the mRNA sequence encoding at least two mRNAs of two or more mammalian hydrolases, such as in the 3' untranslated region (3'UTR), coding region or 5' untranslated region (5'UTR) of at least two mRNAs encoding two or more mammalian hydrolases, preferably in the 3' untranslated region or coding region of at least two mRNAs encoding two or more mammalian hydrolases, more preferably in the coding region of at least two mRNAs encoding two or more mammalian hydrolases, and even more preferably in a microconserved region in the coding region of at least two mRNAs encoding two or more mammalian hydrolases. Typically, the target sequence of all at least two mRNAs encoding two or more mammalian hydrolases is in its 3' untranslated region or coding region, preferably in the coding region of at least two mRNAs encoding two or more mammalian hydrolases, and even more preferably in a microconserved region in the coding region of at least two mRNAs encoding two or more mammalian hydrolases.

[0067] In certain embodiments, the target sequence is located in a conserved region of the amino acid sequence of two or more mammalian hydrolases. In certain preferred embodiments, the target sequence is located in a conserved region of at least two mRNAs encoding two or more mammalian hydrolases, preferably in a conserved region of the coding region of at least two mRNAs encoding two or more mammalian hydrolases. As used herein, the term "conserved region" refers to identical or similar nucleotide sequences in at least two mRNAs encoding two or more mammalian hydrolases. Wherein similar sequences mean a pairwise identity percentage of more than 55%. Pairwise identity refers to the percentage of identical paired residues in the alignment, including gaps relative to non-gap residues, but excluding gaps relative to gap residues. A microconserved region is an identical or similar sequence of a very small segment of at least 21 nucleotides, preferably with at least 55% sequence identity.

[0068] According to the artificial miRNA targeting two or more mammalian hydrolases of the present invention, the activity of the two or more hydrolases described in mammalian cells is reduced. The activity reduction is typically due to the expression caused by mRNA cutting and / or mRNA translation inhibition being suppressed (or prevented). In certain embodiments, the artificial miRNA reduces the hydrolase activity in mammalian cells, such as human or rodent cells, preferably rodent cells, and more preferably CHO cells. In a preferred embodiment, the artificial miRNA reduces the activity of PS-degrading hydrolases and / or prevents the expression of at least two endogenous hydrolases in mammalian cells and / or HCCF obtained from the mammalian cells (such as human or rodent cells). Reduced activity means in this context that compared with mammalian cells not containing artificial miRNA according to the present invention, it is reduced by at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90% or 100%. The hydrolase activity or PS-degrading hydrolase activity in mammalian cells is preferably measured in the cell culture fluid (HCCF) harvested. In addition, the artificial miRNA of the present invention prevents the hydrolase activity and / or polysorbate degrading activity associated with the target recombinant protein produced in the mammalian cell. Preventing expression or preventing hydrolase activity and / or polysorbate degrading activity means any reduction in expression and / or any reduction associated with the target recombinant protein produced in the mammalian cell comprising the artificial miRNA compared to the mammalian cell not comprising the artificial miRNA, such as a reduction of at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90% or 100% compared to the mammalian cell not comprising the artificial miRNA.

[0069] miRNA typically has multiple targets and can regulate multiple genes because its artificial miRNA guide strand sequence can bind to the target sequence (and inhibit expression) via incomplete complementarity. The artificial miRNA guide strand sequence of the artificial miRNA of the present invention typically has 19 to 25 nucleotides, preferably 20 to 24 nucleotides, more preferably 21 to 23 nucleotides, and even more preferably 22 to 23. Binding to one or more target sequences of mRNA requires binding to a 5' seed region and optionally a 3' complementary region (rather than the entire guide strand sequence). Therefore, preferably the artificial miRNA guide strand sequence comprises at least a 5' seed region and a 3' complementary region. In certain embodiments, the artificial miRNA guide strand sequence does not have complete complementarity with at least one of at least two mRNAs encoding two or more mammalian hydrolases, and the miRNA guide strand has a 5' seed region and a 3' complementary region that undergoes base pairing with the one or more target sequences, preferably starting from a 5' seed region of nucleotide 2 (nt 2) at the 5' end of the artificial miRNA guide strand sequence. In general, the artificial miRNA guide strand sequence can have at least 55%, preferably at least 60%, and more preferably at least 65% sequence complementarity with one or more target sequences of at least two mRNAs encoding two or more mammalian hydrolases. In certain embodiments, the miRNA guide strand sequence has at least 13, preferably 14, and more preferably 15 Watson-Crick base pairings (continuous and non-continuous) with one or more target sequences of at least two mRNAs encoding two or more mammalian hydrolases. As used herein, the term "at least 55% sequence complementarity" means that the complementary strand sequence of the miRNA guide strand has at least 55%, preferably at least 60%, and more preferably 65% ​​sequence identity with one or more target sequences, or the antisense transcript of one or more target sequences has at least 55%, preferably at least 60%, and more preferably 65% ​​sequence identity with the miRNA guide strand sequence. In certain embodiments, the content of G and C complementary to the artificial miRNA guide strand sequence and one or more target sequences is 35%-45% (35-45% GC complementarity), preferably 35-42%. Therefore, the 22 to 23 nucleotide artificial miRNA guide strand sequence comprises 8-10, preferably 8-9 G or C that are complementary to (i.e., undergo base pairing with) one or more target sequences. In certain preferred embodiments, the first nucleotide at the 5' end of the artificial miRNA guide strand sequence (i.e., the 5' nucleotide of the 5' seed region) comprises a nucleotide (comprising a nucleobase) selected from adenine (A) and uracil (U), preferably the nucleotide is U.

[0070] In certain embodiments, the artificial miRNA guide strand sequence independently hybridizes with the target sequences of at least two mRNAs encoding two or more mammalian hydrolases without forming a convex loop and / or forming at most two convex loops, preferably without forming a convex loop and / or forming at most one convex loop, more preferably without forming a convex loop. In certain embodiments, the 5' seed region of the artificial miRNA guide strand sequence independently hybridizes with the target sequences of at least two mRNAs encoding two or more mammalian hydrolases without forming a convex loop and / or forming at most one convex loop, more preferably without forming a convex loop. As used herein, the term "convex loop" refers to the following structure in a polynucleotide duplex: one of the chains contains a non-terminal additional sequence that cannot base pair with the second chain, thereby forming a bulge on one side of the duplex.

[0071] The two or more mammalian hydrolases acting on ester bonds may be any mammalian hydrolases, in particular two or more mammalian lipases or carboxylesterases and / or two or more PS degrading mammalian hydrolases. In certain embodiments, at least one of the two or more mammalian hydrolases is selected from lipoprotein lipase (LPL, such as an RNA sequence having SEQ ID NO: 6 or 7), phospholipase B-like protein 2 (PLBL2, such as an RNA sequence having SEQ ID NO: 8 (rna-XM_027414155.1)), phospholipase A1 (Pla1a, such as an RNA sequence having SEQ ID NO: 9 or 10 (rna-XM_027412692.1 or rna-XM_035444560.1, respectively), inactive pancreatic lipase-related protein 1 (PNLIPRP1, such as an RNA sequence having SEQ ID NO: 11 (CgPICR_007696)), inactive pancreatic lipase-related protein 2 (PNLIPRP2; such as an RNA sequence having SEQ ID NO: 12 (rna-XM_027406893.1)), pancreatic triglyceride lipase precursor (PNLIP, such as an RNA sequence having SEQ ID NO: 13 (rna-XM_027406893.1)). ID NO: 13 or 14 RNA sequence (rna-XM_035441745.1 or rna-XM_035441744.1, respectively), lipase member I precursor (LipI, such as the RNA sequence with SEQ ID NO: 15 (rna-XM_035444258.1)), lipase member H precursor (LipH, such as the RNA sequence with SEQ ID NO: 16, 17, 18 or 19 (rna-XM_035444694.1, rna-XM_027413008.1, rna-XM_027413009.1 or rna-XM_027413010.1, respectively), lipase member G precursor (LipG, such as the RNA sequence with SEQ ID NO: NO: 20 RNA sequence (rna-XM_035438714.1)), lipase member C precursor (LipC, such as having SEQ ID NO: 21 RNA sequence (rna-XM_027411221.1)) and their transcript variants or isoforms. Preferably, one of the two or more mammalian hydrolases is lipoprotein lipase (LPL).In certain embodiments, the two or more mammalian hydrolases are selected from lipoprotein lipase (LPL), phospholipase B-like protein 2 (PLBL2), phospholipase A1 (Pla1a), inactive pancreatic lipase-related protein 1 (PNLIPRP1), inactive pancreatic lipase-related protein 2 (PNLIPRP2), pancreatic triacylglycerol lipase precursor (PNLIP), lipase member I precursor (LipI), lipase member H precursor (LipH), lipase member G precursor (LipG), lipase member C precursor (LipC), and transcript variants or isoforms thereof. Preferably, one of the two or more mammalian hydrolases is lipoprotein lipase (LPL).

[0072] All exemplary artificial miRNA guide strand sequences were designed to bind to LPL. The artificial miRNA binding positions are indicated in Table 1 below and in the two LPL sequences.

[0073] Table 1

[0074]

[0075] LPL, lipoprotein lipase (subtype X1), >rna-XM_027388880.1CriGri_PICRH_1.0; (SEQ ID NO: 6)

[0076]

[0077]

[0078] LPL, lipoprotein lipase (subtype X2), >rna-XM_027388881.1CriGri_PICRH_1.0; SEQ ID NO:7)

[0079]

[0080]

[0081] In certain embodiments, one of the two or more mammalian hydrolases is LPL, and the one or more target sequences in the mRNA encoding LPL are within nucleotides 560-590, 700-745, 930-960 or 1330-1360 of SEQ ID NO: 6 or 7, preferably within nucleotides 560-590, 705-745, or 930-960 of SEQ ID NO: 6 or 7. In the context of the present invention, suitable artificial miRNA guide strand sequences are, for example, artificial miRNA guide strand sequences having the nucleotide sequence of SEQ ID NO: 1 (LP 8mer), SEQ ID NO: 2 (LP 6mer), SEQ ID NO: 3 (LH 8mer), SEQ ID NO: 4 (LH 9mer) and / or SEQ ID NO: 5 (non-classical LH).

[0082] The inventors have identified two types of artificial miRNA guide strand sequences that effectively reduce the expression and / or activity of hydrolases. The first sequence type comprises a strong 5' seed region, where strong means high GC content, high complementarity (≥62.5% sequence complementarity, preferably ≥75% sequence complementarity, or no more than 3 mismatches in 8 nucleotides, preferably no more than 2 mismatches in 8 nucleotides); and a weak or normal 3' complementary region (if present) that can compensate for mismatches in the seed region. The second sequence comprises a weaker 5' seed region, which is attributed to a lower GC content or a shorter complementary region, particularly such as having continuous mismatches at the ends of the region; and a strong 3' complementary region that compensates for weaker binding and / or mismatches in the 5' seed region.

[0083] As used herein, the term "5' seed region" refers to nucleotides 2 to 9 of the artificial miRNA guide strand sequence. The miRNA has incomplete complementarity with its target sequence and typically has a higher degree of complementarity in the 5' seed region than the remaining sequence, up to complete complementarity in the 5' seed region. The 5' nucleotide (nucleotide 1) of the 5' seed region is typically A or U and is preferably not complementary to its target sequence.

[0084] As used herein, the term "3' complementary region" refers to the last 6 nucleotides in the artificial miRNA guide strand sequence, also referred to as nucleotides-1 to -6 (nt-1, nt-2, nt-3, nt-4, nt-5, and nt-6), wherein nucleotide-1 is the last nucleotide (or the first nucleotide from the 3' end), and nucleotide-6 is the sixth nucleotide from the 3' end of the artificial miRNA guide strand sequence. Typically, this region has a relatively low GC content, such as a GC content of ≤50%, and usually has two or more mismatches. A higher degree of complementarity (particularly GC complementarity), as well as the last nucleotide being complementary to its target sequence increases the strength of the 3' complementary region.

[0085] In certain embodiments, the miRNA guide strand sequence has incomplete complementarity with one or more target sequences of at least one of at least two mRNAs encoding two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) a sequence of 8 nucleotides (nt 2-9) that is complementary to the one or more target sequences and has 0, 1, 2, or 3 mismatches with the one or more target sequences, and (ii) a GC content of 55-70%, wherein all or all but one or all but two of the Gs or Cs are complementary to the one or more target sequences; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides (i.e., nt-6 to nt-1) that is complementary to the one or more target sequences and has 1, 2, or 3 mismatches with the one or more target sequences, and (ii) a GC content of ≤50% in the last 6 nucleotides, wherein 1 or 2 complementary nucleotides are G or C. Examples of suitable artificial miRNA guide strand sequences are the nucleotide sequences of SEQ ID NO: 2 (LP 6mer), SEQ ID NO: 4 (LH 9mer) or SEQ ID NO: 5 (non-classical LH).

[0086] As used herein, the term "GC content" refers to the percentage of G or C nitrogenous bases in a DNA or RNA molecule (mainly RNA molecules in the context of the present invention). It will be understood by those skilled in the art that 1 or 2 nucleotides are G or C, meaning that 1 or 2 complementary nucleotides contain a nucleobase G or C. Nucleotides are composed of a nucleobase, a pentose (also referred to as a nucleoside) and a phosphate, so for simplicity, nucleotides are defined and referred to by their nucleobases. In addition, nucleobases are referred to by their commonly used abbreviations adenine (A), cytosine (C), guanine (G), thymidine (T) and uracil (U).

[0087] Preferably, the 5'seed region comprising a sequence of 8 nucleotides (nt 2-9) complementary to one or more target sequences has 0, 1 or 2 mismatches, more preferably 0 or 1 mismatches with one or more target sequences. Moreover, the GC content of the 5'seed region is 55-70%, such as 62.5% (or 5 G or C in 8 nucleotides), and preferably all or all G or C except one are complementary to one or more target sequences. In some embodiments, the 5'seed region is longer, such as comprising at least 10 nucleotides (at least nt 2-11) or at least 11 nucleotides (at least nt 2-12), otherwise as described herein.

[0088] In certain embodiments, the 5' seed region as disclosed herein does not comprise mismatches with one or more target sequences within nucleotides 2-3, preferably within nucleotides 2-4, more preferably within nucleotides 2-5, and even more preferably within nucleotides 2-7, and / or the 5' seed sequence comprises at least 2 3-4 consecutive base pairings with one or more target sequences. In a preferred embodiment, the one or more mismatches with one or more target sequences in the 5' seed region are not consecutive mismatches. In certain embodiments, the first nucleotide at the 5' end of the artificial miRNA guide strand sequence (i.e., the 5' nucleotide of the 5' seed region) comprises a nucleotide (comprising a nucleobase) selected from A and U, preferably the nucleotide is U.

[0089] As disclosed herein, the 3' complementary region comprises a sequence of the last 6 nucleotides (i.e., nt-6 to nt-1) complementary to one or more target sequences, with no more than three mismatches in the last 6 nucleotides of the miRNA guide strand and the target sequence, such as 1, 2 or 3 mismatches with one or more target sequences, preferably 1 or 2 mismatches with one or more target sequences. In a preferred embodiment, the mismatches in the 3' complementary region are not continuous mismatches. Moreover, the 3' complementary region has a GC content of ≤50% in the last 6 nucleotides, such as 1, 2 or 3 G or C, preferably 2 or 3 G or C (33% to 50% GC content), wherein 1 or 2 complementary nucleotides are G or C. It will be appreciated by those skilled in the art that increasing the number of mismatches and / or continuous mismatches and / or a lower number of G or C complementary nucleotides makes the 3' complementary region weaker, while reducing the number of mismatches and / or no continuous mismatches and / or a higher number of G or C complementary nucleotides makes the 3' complementary region stronger. Furthermore, a stronger 3' complementary region can potentially compensate for a weaker 5' seed region.

[0090] In certain alternative embodiments, the miRNA guide strand sequence has incomplete complementarity with one or more target sequences of at least one of at least two mRNAs encoding two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) a sequence of 8 nucleotides (nt 2-9), wherein at least the first 4 consecutive nucleotides (at least nt 2-5) are complementary to the one or more target sequences, wherein the sequence of 8 nucleotides comprises at least one mismatch with the one or more target sequences, and (ii) a GC content of 55-75%, wherein at least 4 consecutive nucleotides complementary to the one or more target sequences are G or C; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides that are complementary to the one or more target sequences and have 0 or 1 mismatches with the one or more target sequences, and (ii) a GC content of 30-50% in the last 6 nucleotides, preferably wherein at least 2 complementary nucleotides are G or C. An example of a suitable artificial miRNA guide strand sequence is the nucleotide sequence of SEQ ID NO: 1 (LP 8mer) or SEQ ID NO: 2 (LP 6mer).

[0091] According to alternative embodiments, the 5'seed region comprises a sequence of 8 nucleotides (nt 2-9), wherein at least the first 4 consecutive nucleotides (at least nt 2-5) are complementary to one or more target sequences, wherein the sequence of 8 nucleotides comprises at least one mismatch with one or more target sequences. Therefore, at least one mismatch is in nucleotides 6-9, and at least one mismatch can be 1, 2, 3 or 4 mismatches. In certain embodiments, the 5'seed region comprises a sequence of 8 nucleotides (nt 2-9), wherein at least the first 6 consecutive nucleotides (at least nt 2-7) are complementary to one or more target sequences, wherein the sequence of 8 nucleotides comprises at least one mismatch with one or more target sequences. Therefore, at least one mismatch is in nucleotides 8-9, and at least one mismatch can be 1 or 2 mismatches. As disclosed herein, the GC content in the 5'seed region is 55-75% (such as 62.5 to 75%, or as 5 or 6 G or C in 8 nucleotides), wherein at least 4 consecutive nucleotides complementary to one or more target sequences are G or C. For example, 5 or 6 consecutive nucleotides complementary to one or more target sequences are G or C. Preferably, at least the first 4 consecutive nucleotides are complementary to one or more target sequences and are G or C. In a specific embodiment, at least the first 4 consecutive nucleotides (nt 2-5) are complementary to one or more target sequences and are G or C, and the 8-nucleotide sequence comprises at least two mismatches with one or more target sequences in nucleotides 6-9, preferably 3 or 4 mismatches in nucleotides 6-9. In certain embodiments, the first nucleotide at the 5' end of the artificial miRNA guide strand sequence (i.e., the 5' nucleotide of the 5' seed region) comprises a nucleotide selected from A and U (comprising a nucleobase), preferably the nucleotide is U.

[0092] As disclosed herein, the corresponding 3' complementary region comprises (i) a sequence of the last 6 nucleotides, which is complementary to one or more target sequences and has 0 or 1 mismatches with one or more target sequences. Preferably, the last nucleotide (nt-1) is complementary to one or more target sequences, and / or the penultimate nucleotide is a mismatch, more preferably the last nucleotide is complementary to one or more target sequences and the penultimate nucleotide is a mismatch. As disclosed herein, the GC content in the 3' complementary region is 30-50% (such as 33% to 50%, or such as 2 or 3 G or C in 8 nucleotides) in the last 6 nucleotides, wherein at least 2 complementary nucleotides are G or C, preferably wherein 2 complementary nucleotides are G or C. As used herein, the term complementary nucleotide means complementary to one or more target sequences.

[0093] The artificial miRNA guide strand sequence according to the present invention is designed from scratch, which means that it is not naturally present in mammalian cells or organisms. Therefore, the sequence is generated by a rational design based on the sequence alignment of the target sequence at the mRNA and / or protein level (preferably at the mRNA level). The object of the present invention is to provide the following proof of concept: the artificial miRNA comprising the artificial miRNA guide strand sequence can be used for cell line engineering methods, for simultaneously targeting multiple hydrolases acting on ester bonds in mammalian host cells such as CHO cells, such as multiple lipases. The purpose is to reduce the hydrolase expression in mammalian host cells, HCCF and the purified purpose recombinant protein produced in the mammalian host cells, and thus reduce hydrolase activity, PS degradation enzyme activity, PS degradation and / or particle formation. In particular, the purpose is to reduce the expression of two or more hydrolases (such as lipases and / or carboxylesterases) acting on ester bonds, so as to reduce the presence and / or PS degradation activity of two or more hydrolases acting on ester bonds in the purified purpose recombinant protein produced in the mammalian host cells and / or HCCF and / or the mammalian host cells, or in its composition. In addition, it is reported that some lipases such as PLBL2 are immunogenic in patients. Therefore, reducing the expression of two or more hydrolases (such as lipases and / or carboxylesterases) acting on ester bonds in mammalian host cells can also lead to a reduction in the immunogenic activity of the purified recombinant protein of interest or its composition produced in mammalian host cells. It has been observed that the hydrolytic activity spectrum co-purified with the target protein varies with biopharmaceuticals, so contaminating co-purified hydrolases (lipases and carboxylesterases) are considered to be different between target proteins, cell lines, culture conditions and cell clones. Therefore, methods targeting multiple targets will be helpful.

[0094] The highly conserved transcripts of lipases are therefore targeted by de novo designed artificial microRNAs (amiRNAs). As proof of concept, artificial (de novo designed) miRNAs have been shown to simultaneously knock down lipases via incomplete mRNA target binding, so artificial miRNAs containing artificial (de novo designed) miRNA guide strand sequences can be used to stably express with a transgene of interest to target and reduce the expression of multiple (two or more) hydrolases acting on ester bonds (such as lipases or carboxylesterases) in mammalian host cells.

[0095] The artificial miRNA according to the present invention can be pri-miRNA, pre-miRNA, double-stranded mature miRNA (comprising a follower strand and a guide strand) or a single-stranded miRNA guide strand. In certain embodiments, the artificial miRNA is a single-stranded or double-stranded miRNA mimic (miR-Mimic). MiRNA mimics are chemically synthesized double-stranded RNAs that simulate mature miRNAs after transfection into cells, and can be provided as double-stranded miRNA mimics or as single-stranded miRNAs, the latter requiring an additional annealing step with the follower strand prior to transfection. Customized miRNA mimics can be obtained by proprietary chemical modifications in order to improve their performance and stability, and to prevent the binding activity of follower strands obtained on demand from different suppliers (such as from Horizon, Qiagen or Thermo Fisher).

[0096] Expression vector encoding artificial miRNA targeting two or more hydrolases

[0097] In another aspect, the present invention relates to a mammalian expression vector comprising a polynucleotide sequence encoding an artificial miRNA according to the present invention. The mammalian expression vector may also comprise at least one target gene encoding a target recombinant protein. Alternatively, the target recombinant protein may also be encoded by a separate expression vector.

[0098] In certain embodiments, the mammalian expression vector comprises a polynucleotide sequence encoding two or more artificial miRNAs according to the present invention, wherein the artificial miRNAs are the same or different. The two miRNAs may be encoded by the same or separate expression cassettes.

[0099] The mammalian expression vector according to the invention is used to express heterologous sequences in mammalian cells, i.e., it is suitable for expressing in mammalian cells, i.e., an artificial miRNA and optionally an additional protein of interest. Therefore, the mammalian expression vector according to the invention is characterized in that it comprises a polynucleotide operably linked to a mammalian promoter, the polynucleotide sequence encoding a polynucleotide sequence encoding an artificial miRNA. Typically, the mammalian expression vector comprises a mammalian cassette comprising a polynucleotide sequence encoding an artificial miRNA operably linked to a mammalian promoter. The mammalian promoter regulates transcription in mammalian cells. Exemplary mammalian promoters are, but are not limited to, the simian virus 40 (SV40) early promoter, the cytomegalovirus (CMV) immediate early promoter, the Rous sarcoma virus (RSV) promoter, adenovirus promoters (e.g., adenovirus major late promoter (AdMLP), human ubiquitin C promoter (UBC), human elongation factor 1 alpha promoter (EF1A), CHO-derived elongation factor 1 (CHEF-1) mouse phosphoglycerate kinase 1 promoter (PGK), and chicken beta actin promoter coupled to CMV early enhancer (CAGG), as well as strong mammalian promoters such as natural immunoglobulin and actin promoters, or the native promoter of a heterologous polynucleotide (e.g., a polynucleotide encoding a protein of interest). Mammalian expression vectors may further comprise bacterial sequences, such as replication origins and resistance genes for vector amplification in bacterial cells.

[0100] As used herein, the term "expression cassette" is a unique component of DNA (particularly vector DNA) composed of one or more coding polynucleotide sequences and regulatory sequences that control their expression in transfected or transduced cells. The expression cassette contains at least three components: a promoter sequence, an open reading frame (such as a polynucleotide encoding a target protein) and a termination sequence. It may further contain means for controlling the expression of one or more gene products (mRNA encoding a target protein), such as an enhancer. In a mammalian expression vector comprising a polynucleotide encoding a target protein, the termination sequence is called 3'UTR and generally contains a polyadenylation site. The expression cassette instructs the cell's machinery to make RNA, and therefore may also be referred to as a transcription cassette. MiRNA can be expressed in an expression cassette of a target recombinant protein or preferably a selective marker (i.e., in an intron), from which pre-miRNA is spliced ​​out and processed until mature miRNA is generated by the cell machinery. Alternatively, it can be expressed separately on the same expression vector or on a separate expression vector and driven by its own promoter.

[0101] Exemplary mammalian promoters are, but are not limited to, the simian virus 40 (SV40) early promoter, the cytomegalovirus (CMV) immediate early promoter, the Rous sarcoma virus (RSV) promoter, adenovirus promoters (e.g., adenovirus major late promoter (AdMLP), human ubiquitin C promoter (UBC), human elongation factor 1 alpha promoter (EF1A), CHO-derived elongation factor 1 (CHEF-1) mouse phosphoglycerate kinase 1 promoter (PGK), and chicken beta actin promoter coupled to CMV early enhancer (CAGG), as well as strong mammalian promoters such as natural immunoglobulin and actin promoters, or the native promoter of a heterologous polynucleotide (e.g., a polynucleotide encoding a protein of interest). Examples of polyadenylation signals are BGH polyA, SV40 late or early polyA; alternatively, the 3'UTR of an immunoglobulin gene or the like can be used. It will be further appreciated by those skilled in the art that the 3'UTR can be engineered to support high expression levels, for example, by removing unstable elements such as ARE (adenylate-uridinate-rich element).

[0102] The target recombinant protein can be any protein, but is typically a therapeutic protein. As used herein, the term "therapeutic protein" refers to a protein that can be used for medical treatment of humans or animals. These include, but are not limited to, cytokines, growth factors, hormones, coagulation factors, vaccines, interferons, fusion proteins, antibodies, antibody-derived molecules, and mimetic antibodies. In certain embodiments, the therapeutic protein is selected from cytokines, hormones, fusion proteins, antibodies, antibody-derived molecules, and mimetic antibodies. As used herein, the term "target gene" refers to a polynucleotide sequence (DNA sequence) encoding a target recombinant protein.

[0103] In some embodiments, the gene product (miRNA or target protein) can be placed under the control of an amplifiable genetic selection marker such as dihydrofolate reductase (DHFR) or glutamine synthetase (GS). The amplifiable selection marker gene can be on the same expression vector as the target gene and / or the polynucleotide sequence encoding the artificial miRNA of the present invention. Alternatively, the amplifiable selection marker gene and the secretory therapeutic protein expression cassette can be on different expression vectors, but tightly integrated into the genome of the host cell. For example, two or more vectors co-transfected at the same time are usually tightly integrated into the genome of the host cell. In addition, the amplification of the genetic region containing the secretory therapeutic protein expression cassette can be mediated by adding an amplifying agent (e.g., methotrexate (MTX) for DHFR or methionine sulfoximide (MSX) for GS) to the culture medium.

[0104] Sufficiently high stable levels of gene products can be achieved in host cells or production cells, for example, by cloning multiple copies of polynucleotides encoding artificial miRNAs and / or target proteins of the present invention into expression vectors. Multiple copies of polynucleotides can be cloned into expression vectors and expression cassettes (encoding artificial miRNAs or target recombinant proteins) can be amplified for further combination.

[0105] Mammalian expression vectors may include, but are not limited to, plasmid vectors, cosmids, artificial / mini-chromosomes (e.g., ACE), bacterial artificial chromosomes (BACs), or viral vectors, such as lentivirus, retrovirus, adenovirus, adeno-associated virus, herpes simplex virus, and retrovirus-derived vectors. Preferably, the mammalian expression vector is a plasmid vector, a bacterial artificial chromosome (BAC), or a viral vector (e.g., a lentiviral vector). The mammalian expression vector can be introduced into a mammalian host cell via transfection or transduction, respectively, and is preferably stably integrated into the host cell genome. Those skilled in the art know suitable plasmids, BACs, or viral vectors, and plasmids may further include elements for targeted integration (homologous regions) upstream and downstream of the polynucleotide sequence encoding artificial miRNA and / or target protein (and optionally selective markers such as GS), including transposon recognition sequences and reverse terminal repeats.

[0106] Means for cloning microRNAs into expression vectors are known to those skilled in the art. For example, one or more microRNAs can be cloned into a mammalian expression vector as a polynucleotide encoding an engineered pre-miRNA sequence (i.e., a short hairpin), such as pcDNA6.2-GW / miR or pcDNA6.2-GW / EmGFP-miR from Invitrogen (see the manual BLOCK-iT TMPol II miR RNAi expression vector kit). The vector can encode one or more copies of the same or different miRNAs. In short, the mature miRNA sequence is cloned into a given sequence encoding an optimized hairpin loop sequence and 3' and 5' flanking regions derived from mouse miRNA mir-155 (Lagos-Quintana et al., 2002). The flanking regions are present on the vector, and a DNA oligonucleotide is designed that encodes the miRNA guide strand sequence, the mentioned loop and antisense sequence of the corresponding mature miRNA, wherein two nucleotides are deleted to generate an inner loop in the hairpin stem. In addition, overhangs are added at both ends for cloning. The hairpin structure can be analyzed using the online tool mfold (M.Zuker, Mfold web server fornucleic acid folding and hybridization prediction, Nucleic Acids Res.31 (13), 3406-3415, 2003). As described by the manufacturer, the DNA strand is annealed and connected to the 3'UTR of the emerald GFP reporter protein gene. The application of the linkage method can generate vectors containing more than one miRNA. Negative control miRNA (supplied by the manufacturer) and siLacZ can be used as appropriate negative controls. Alternative vectors that can be used for miRNA expression in the present invention are but are not limited to pCMV-MIR (Origene), pmR-ZsGreen1 (clontech) and shMIMIC lentiviral miRNA vectors (ThermoScientific).

[0107] The target recombinant protein can be any protein, and is typically a therapeutic protein. As used herein, the term "therapeutic protein" refers to a protein that can be used for medical treatment of humans or animals. These include, but are not limited to, cytokines, growth factors, hormones, coagulation factors, vaccines, interferons, fusion proteins, antibodies, antibody-derived molecules, and mimetic antibodies. In certain embodiments, the therapeutic protein is selected from cytokines, hormones, fusion proteins, antibodies, antibody-derived molecules, and mimetic antibodies.

[0108] In certain embodiments, the target protein is an antibody. In the case where the target protein is an antibody, the mammalian expression vector comprises a polynucleotide comprising a coding sequence for a heavy chain variable region and / or a coding sequence for a light chain variable region of an antibody. In certain embodiments, the eukaryotic expression vector comprises a polynucleotide comprising a coding sequence for a heavy chain and / or a coding sequence for a light chain of an antibody. Therefore, the polynucleotide comprising the heavy chain variable region coding sequence and the polynucleotide comprising the light chain variable region coding sequence can be expressed on the same mammalian expression vector or on a separate mammalian expression vector. The expression vector may comprise a polycistronic expression cassette such as a bicistronic expression cassette, and / or a plurality of expression cassettes. The polycistronic expression cassette comprises more than one open reading frame, which is separated by sequences encoding RNA elements (such as internal ribosome entry sites (IRES)) that allow translation initiation. In a polycistronic expression cassette, two or more open reading frames are under the control of the same promoter. Thus, the polynucleotide encoding at least the heavy chain variable region and the polynucleotide encoding at least the light chain variable region can be expressed in the same expression cassette (eg, separated by an IRES sequence) or by two separate expression cassettes.

[0109] Mammalian cells comprising artificial miRNAs targeting two or more hydrolases

[0110] In another aspect, the present invention relates to a mammalian cell (preferably a CHO cell) comprising an artificial miRNA of the present invention or a polynucleotide encoding an artificial miRNA of the present invention or a mammalian expression vector of the present invention. The mammalian cell may be any mammalian cell suitable for expressing proteins in cell culture, such as a human cell or a rodent cell. Preferably, the mammalian cell is a CHO cell.

[0111] Compared to mammalian cells that do not contain the artificial miRNA of the present invention, mammalian cells that contain the artificial miRNA of the present invention have reduced expression of two or more mammalian hydrolases. In certain embodiments, mammalian cells that contain the artificial miRNA of the present invention and / or the target recombinant protein produced and separated from the mammalian cells have reduced hydrolase activity and / or PS degradation activity compared to mammalian cells that do not contain the artificial miRNA of the present invention and / or the target recombinant protein produced and separated from the mammalian cells. Mammalian cells that contain the artificial miRNA of the present invention and / or the target recombinant protein produced and separated from the mammalian cells may also have reduced PS degradation activity of two or more mammalian hydrolases compared to mammalian cells that do not contain the artificial miRNA of the present invention and / or the target recombinant protein produced and separated from the mammalian cells. It should be understood by those skilled in the art that the recombinant protein produced and separated from the mammalian cells includes a fluid or composition containing the recombinant protein, such as HCCF, a purified or partially purified recombinant protein, or a purified and formulated composition (bulk drug or drug). "Isolated" includes purification and separation, such as by centrifugation or filtration of the cell culture supernatant. In certain embodiments, artificial miRNA prevents the expression of at least two endogenous hydrolases in mammalian cells, and / or prevents the hydrolase activity and / or polysorbate degradation activity associated with the target recombinant protein produced in mammalian cells. Being associated with the target recombinant protein produced in mammalian cells means being co-purified with the target recombinant protein produced in mammalian cells and / or in the composition comprising the target recombinant protein produced in mammalian cells. It can be by means of, for example, WO 2022 / 049294 for detecting lipase activity using 4-methylumbelliferyldecanoate (4-MUD) as a substrate, and / or in EP 22161135 for detecting hydrolases (such as carboxylesterases) using 1-octanoyloxy-pyrene-3,6,8-trisulfonic acid (OPTS) as a substrate, the fluorometric method described (but not limited thereto) to determine the reduced hydrolase activity. PS concentration can be measured, for example, using fluorescent micelle determination (FMA) as described herein. Therefore, to determine the reduced PS degradation activity, PS20 (e.g., cPS20 = 0.4 mg / ml) or PS80 (e.g., cPS80 = 0.2 mg / ml) can be added to the sample, such as the HCCF sample (diluted), and then stored at room temperature (about 22 ° C) for 0, 1, 3, 7 and 14 days, and frozen at -70 ° C before analysis using FMA. For FMA, 10 μl of sample was mixed with 240 μl of FMA reagent (150 mM NaCl, 50 mM Tris, 0.2% acetonitrile, 5 μM NPN, 0.0015% Brij-35, pH 8), followed by incubation at 35 ° C for 1 minute. The sample was analyzed using a microplate reader (such as obtained from 200PRO Tecan, Switzerland) after exciting the dye with a 350nm laser, a fluorescence signal positively correlated with the PS concentration can be measured at 420nm. The PS concentration is quantified using a standard curve of PS20 (e.g., in the range of 0.1 to 0.6mg / ml) and / or PS80 (e.g., in the range of 0.05 to 0.3mg / ml). The reduction compared to the control (a sample obtained from cells without artificial miRNA) can be determined by comparing PS20 or PS80 (in mg / ml) on different days or by comparing the degradation rate (mg / ml / day).

[0112] The mammalian cell according to the present invention can also express a target recombinant protein, preferably a therapeutic protein, more preferably a therapeutic antibody. The target recombinant protein can be stably or transiently expressed, preferably the target recombinant protein is stably expressed.

[0113] In certain embodiments, mammalian cells according to the present invention are transfected or transduced with a mammalian expression vector of the present invention. Preferably, mammalian cells are stably transfected or transduced with a mammalian expression vector of the present invention. Methods for transfecting or transducing mammalian cells are well known in the art and include semi-targeted, targeted, and random integration. Alternatively, mammalian cells according to the present invention are transfected with an artificial miRNA of the present invention or with a mammalian expression vector comprising a polynucleotide sequence encoding an artificial miRNA of the present invention. Methods for transfecting mammalian cells with miRNA or with a mammalian expression vector are well known in the art.

[0114] The target recombinant protein can be any protein, and is typically a therapeutic protein. As used herein, the term "therapeutic protein" refers to a protein that can be used for medical treatment of humans or animals. These include, but are not limited to, cytokines, growth factors, hormones, coagulation factors, vaccines, interferons, fusion proteins, antibodies, antibody-derived molecules, and mimetic antibodies. In certain embodiments, the therapeutic protein is selected from cytokines, hormones, fusion proteins, antibodies, antibody-derived molecules, and mimetic antibodies.

[0115] The preferred recombinant protein of interest is an antibody, including fragments and derivatives thereof. Typically, antibodies are monospecific, but antibodies may also be multispecific (symmetrical or asymmetrical). Therefore, the present invention can be used to produce monospecific antibodies, multispecific antibodies, or fragments thereof, preferably antibodies (monospecific), bispecific antibodies, trispecific antibodies, or fragments 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-TNFα, anti-IL2, anti-IL-5R or anti-IgE antibodies, and are preferably selected from anti-CD20, anti-CD33, anti-CD37, anti-CD40, anti-CD44, anti-CD52, anti-HER2 / neu (erbB2), anti-EGFR, anti-IGF, anti-VEGF, anti-TNFα, anti-IL2 and anti-IgE antibodies.

[0116] 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), and antibody fragments, as 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 IgG1 or IgG4 antibody.

[0117] Antibodies can be of any kind 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 regions of the heavy and light chains can be derived from rat or mouse antibodies, while the constant regions are derived from human antibodies. In "humanized" antibodies, only a very small sequence is derived from a non-human species. Usually, only the complementary determining region (CDR) amino acid residues of human antibodies are replaced by CDR amino acid residues of non-human species (such as mice, rats, rabbits or llamas). Sometimes, several key framework amino acid residues that have an impact on antigen binding specificity and affinity are also replaced by non-human amino acid residues.

[0118] Typically, an antibody is a tetrameric polypeptide consisting of two pairs of heterodimers, each pair of heterodimers being formed by a heavy chain and a light chain. Stabilization of the heterodimer and tetrameric polypeptide structures occurs via interchain disulfide bonds. Each chain is composed of a domain called an "immunoglobulin domain" or "immunoglobulin region", so the terms "domain" or "region" are used interchangeably. Each domain contains about 70-110 amino acids and forms a compact three-dimensional structure. Both the heavy chain and the light chain contain a "variable domain" or "variable region" whose sequence is not very conservative at their N-termini, 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 is referred to as "VH".

[0119] "Antibody fragment" or "antigen-binding fragment" refers to a molecule other than an intact antibody that contains a portion of an intact antibody that binds to 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 region fragment antibody molecules (e.g., scFv); and multispecific antibodies formed from antibody fragments. Fab fragments consist of the variable regions of two chains held together by adjacent constant regions. These can be formed from conventional antibodies by protease digestion (e.g., with papain), but similarly Fab fragments can also be produced by genetic engineering. Additional antibody fragments include F(ab')2 fragments, which can be prepared by proteolytic cleavage with pepsin.

[0120] Using genetic engineering methods, it is possible to produce shortened antibody fragments consisting only of the variable region of the heavy chain (VH) and the variable region of the light chain (VL). These are called Fv fragments (variable fragment = fragment of the variable part). Since these Fv fragments lack the covalent bonding of the cysteine ​​of the constant chain to the two chains, Fv fragments are usually stabilized. Advantageously, the variable regions of the heavy chain and the light chain are connected by short peptide fragments of, for example, 10 to 30 amino acids, preferably 15 amino acids. In this way, a single peptide chain consisting of VH and VL connected by a peptide linker is obtained. This type of antibody protein is called scFv. Examples of scFv antibody proteins are known to those skilled in the art. Therefore, antibody fragments and antigen-binding fragments also include Fv fragments and particularly scFv.

[0121] In recent years, various strategies have been developed to prepare scFv as a multimer derivative. This is particularly intended to result in recombinant antibodies with improved pharmacokinetic and biodistribution properties and increased binding affinity. In order to achieve the multimerization of scFv, scFv is prepared as a fusion protein with a multimerization domain. The multimerization domain can be, for example, the CH3 (CH3=third constant heavy chain domain) region of IgG or a coiled coil structure (helical structure) such as a leucine zipper domain. However, there are also strategies (e.g., double antibodies, three antibodies and five antibodies) for multimerization using the interaction between the VH / VL regions of scFv. Double antibodies referred to by those skilled in the art mean divalent homodimer scFv derivatives. The shortening of the linker in the scFv molecule to 5-10 amino acids leads to the formation of homodimers, in which interchain VH / VL overlap occurs. Double antibodies can be additionally stabilized by introducing disulfide bridges. Examples of double antibody-antibody proteins are known in the prior art.

[0122] By minibodies, those skilled in the art mean bivalent homodimeric scFv derivatives. They consist of a fusion protein containing the CH3 region of an immunoglobulin (preferably IgG, most preferably IgG1) as a dimerization region, which is connected to the scFv via a hinge region (e.g., also derived from IgG1) and a linker region. Examples of minibodies-antibody proteins are known in the prior art.

[0123] The term triabodies by those skilled in the art refers to trivalent homotrimeric scFv derivatives, wherein VH-VL are directly fused without a linker sequence, resulting in the formation of trimers.

[0124] Those skilled in the art will also be familiar with so-called minibodies, which have a bivalent, trivalent or tetravalent structure and are derived from scFv. Multimerization is carried out through a dimer, trimer or tetramer coiled-coil structure. In a preferred embodiment of the present invention, the target gene encodes any of those desired polypeptides mentioned above, preferably a monoclonal antibody, a derivative or a fragment thereof.

[0125] Further encompassed are single domain antibodies (sdAbs), also known as nanobodies, which are antibody fragments of a single monomeric variable antibody domain. Single domain antibodies are typically produced by camelids (V H H segment) or cartilaginous fish (V NAR The heavy chain antibody present in the fragment is engineered.

[0126] The immunoglobulin fragments consisting of the CH2 and CH3 domains of the antibody heavy chains are called "Fc fragments", "Fc regions" or "Fc" because of their tendency to crystallize (Fc = fragment crystallizable). These fragments can be formed from conventional antibodies by protease digestion (e.g., with papain or pepsin), but can also be produced by genetic engineering. The N-terminal part of the Fc fragment may differ, depending on how many amino acids are still present in the hinge region.

[0127] An antibody comprising an antigen binding fragment and an Fc region may also be referred to as a full-length antibody. Full-length antibodies may be monospecific and multispecific antibodies. Multispecific antibodies are antibodies having at least two different antigen binding sites, each of which binds to a different epitope. Multispecific antibodies include, but are not limited to, bispecific and trispecific antibodies. Bispecific antibodies have two different binding sites. Multispecific antibodies also include antibody forms other than full-length antibodies, such as antibody-derived molecules.

[0128] Bispecific antibodies typically combine antigen binding specificity 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 are, but are not limited to, symmetrical antibodies, double antibodies, BiTE (bispecific T cell engagers) forms, and DART (dual affinity retargeting) forms fused with additional scFv for a second target at the Fc portion. The double antibody format separates the homologous variable domains of the heavy and light chains of the two antigen binding specificities on two separate polypeptide chains, which are non-covalently associated. The DART format is based on the double antibody format, but it provides additional stabilization through a C-terminal disulfide bridge. Trispecific antibodies are monoclonal antibodies that combine three antigen binding specificities. They can be based on bispecific antibody technology, which reconfigures the antigen recognition domains of two different antibodies into a bispecific molecule; or they can be based on a DART structure with scFv at the Fc portion of one chain. For example, trispecific antibodies targeting CD38 on cancer cells and CD3 and CD28 on T cells have been generated. Multispecific (bispecific and trispecific) antibodies are particularly difficult to produce with high product quality.

[0129] As used herein, the term "antibody-derived molecule" refers to any molecule that contains at least an antigen-binding portion that is structurally related to an antibody. It includes modified full-length monospecific or bispecific antibodies that are further modified with additional antigen-binding portions or smaller antibody forms, including those described herein.

[0130] As used herein, the term "mimetic antibody" refers to a protein that binds to a specific antigen in a manner similar to an antibody but is structurally unrelated to an antibody. Mimetic antibodies include, but are not limited to, anticalins, affibodies, adnectins, monoclonal antibodies, DARPins, affimers, and affitins.

[0131] Single domain antibodies (sdAbs) may also be referred to as nanobodies. It will be appreciated by those skilled in the art that a protein may comprise more than one antigen binding domain and may therefore be multivalent, preferably bivalent (eg, a bivalent sdAb or a bivalent anticalin or any other bivalent mimetibody).

[0132] Another preferred therapeutic protein is a fusion protein, such as Fc-fusion protein. Therefore, the present invention can be advantageously used to produce fusion proteins, such as Fc-fusion proteins. The effector portion of the fusion protein can be the complete sequence of a natural or modified heterologous protein or any part of the sequence. The immunoglobulin constant domain sequence can be obtained from any immunoglobulin subtype such as IgG1, IgG2, IgG3, IgG4, IgA1 or IgA2 subtype or classification such as IgA, IgE, IgD or IgM. Preferentially, they are derived from human immunoglobulin, more preferably human IgG, and even more preferably human IgG1 and IgG2. The limiting examples of Fc fusion proteins are MCP1-Fc, ICAM-Fc, EPO-Fc and scFv fragments etc. coupled with the CH2 domains of the heavy chain immunoglobulin constant region of the glycosylation site that comprises N-connection. Fc fusion proteins can be constructed by genetic engineering methods in the following manner: the CH2 domain of the heavy chain immunoglobulin constant region containing an N-linked glycosylation site is introduced into another expression construct containing, for example, other immunoglobulin domains, enzymatically active protein portions, or effector domains. Therefore, the Fc fusion protein according to the present invention also comprises a single-chain Fv fragment connected to the CH2 domain of the heavy chain immunoglobulin constant region containing, for example, an N-linked glycosylation site.

[0133] The term "cytokine" refers to a small protein that is released by a cell and acts as an intercellular mediator (e.g., affects the behavior of cells around the secreting cell). Cytokines can be secreted by immune cells or other cells such as T-cells, B-cells, NK cells and macrophages. Cytokines can participate in intercellular signal transduction events, such as autocrine signal transduction, paracrine signal transduction and endocrine signal transduction. They can mediate a series of biological processes, including but not limited to immunity, inflammation and hematopoiesis. Cytokines can be chemokines, interferons, interleukins, lymphokines or tumor necrosis factors.

[0134] As used herein, "growth factor" refers to a protein or polypeptide that is capable of stimulating cell growth.

[0135] The artificial miRNA, the polynucleotide encoding the artificial miRNA of the present invention, or the expression vector is introduced into mammalian cells for producing the target protein.

[0136] Preferred mammalian cells for producing heterologous or recombinant proteins are rodent cells or human cells. Thus, mammalian cells (also referred to as mammalian host cells) are preferably human or rodent cells, more preferably rodent cells (e.g., mouse or hamster cells), even more preferably hamster cells, such as CHO cells. Suitable rodent cells may be, for example, hamster cells, in particular BHK21, BHK TK-, CHO, CHO-K1, CHO-DXB11 (also referred to as CHO-DUKX or DuxB11), CHO-S cells and CHO-DG44 cells or derivatives / progeny of any such cell lines. Particularly preferred are CHO cells, such as CHO-DG44, CHO-K1 and BHK21, and even more preferred are CHO-DG44 and CHO-K1 cells. Mammalian cells, in particular GS-deficient derivatives of CHO-DG44 and CHO-K1 cells are also contemplated. In one embodiment of the present invention, the mammalian cell is a 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. Suitable human cells are HEK293 or HEK293T cells. The host cell may also be a mouse cell such as a mouse myeloma cell, such as NS0 and Sp2 / 0 cells or a derivative / progeny of any such cell line.

[0137] Preferred examples of mammalian cells or mammalian cell lines are CHO cells (such as DG44 and K1), NS0 cells, HEK293 cells (such as HEK293 cells and HEK293T cells) and BHK21 cells. Preferably, the mammalian cells or mammalian cell lines are suitable for growth in suspension. In a preferred embodiment, the mammalian cells or mammalian cell lines are CHO cells. In certain embodiments, the mammalian cells are HEK293 cells, or CHO cells, or cells derived from HEK293 cells or CHO cells, preferably the mammalian cells are CHO cells or CHO derived cells.

[0138] Preferably, the CHO cells that allow the development of efficient cell lines are metabolically engineered, such as by knocking out GS and / or DHFR to be respectively conducive to selection with MSX or MTX. The conventional CHO cells used for large-scale industrial production are usually engineered to improve their characteristics in the production process or to facilitate the selection of recombinant cells. Such engineering includes but is not limited to increasing apoptosis resistance, reducing autophagy, increasing cell proliferation, changing the expression of cell cycle regulatory proteins, molecular chaperone engineering, unfolded protein response (UPR) engineering, secretory pathway engineering and metabolic engineering.

[0139] Non-limiting examples of mammalian cells that can be used in the context of the present invention are also summarized in Table 2. However, derivatives / progeny of those cells, other mammalian cells (including but not limited to human, mouse, rat, monkey and rodent cell lines) can also be used in the present invention, particularly for the production of biopharmaceutical proteins.

[0140] Table 2: Exemplary mammalian production cell lines

[0141]

[0142]

[0143] 1 CAP (Amniotic Cell Production of CEVEC) cells are an immortalized cell line based on primary human amniotic cells. They are generated by transfecting these primary cells with a vector containing functional E1 and pIX of adenovirus type 5. As a result of authentic human post-translational modifications, CAP cells allow competitive and stable production of recombinant proteins with excellent biological activity and therapeutic efficacy.

[0144] When established, adapted and fully cultured under serum-free conditions and optionally in a culture medium that does not contain any animal-derived protein / peptide and / or chemical composition, cells are most preferred. Commercially available culture media, such as Ham's F12 (Sigma, Deisenhofen, Germany), RPMI-1640 (Sigma), Dulbecco's modified Eagle medium (DMEM; Sigma), minimal essential medium (MEM; Sigma), Iscove's modified Dulbecco's medium (IMDM; Sigma), CD-CHO (Invitrogen, Carlsbad, CA), serum-free CHO culture medium (Sigma), and protein-free CHO culture medium (Sigma) are exemplary appropriate nutrient solutions. Any culture medium can be supplemented with various compounds as needed, and its non-limiting examples 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 known to those skilled in the art.In order to grow and select genetically modified cells expressing the selectable gene, a suitable selection agent is added to the culture medium.

[0145] Use of artificial miRNAs targeting two or more hydrolases

[0146] In another aspect, the present invention relates to the use of a mammalian cell according to the present invention for producing a recombinant protein of interest, preferably a therapeutic protein, more preferably a therapeutic antibody.

[0147] In another aspect, the present invention relates to the use of an artificial miRNA according to the present invention in a mammalian cell for preventing the expression of at least two endogenous hydrolases in a mammalian cell and / or preventing the hydrolase activity and / or polysorbate degradation activity associated with a recombinant protein of interest produced in a mammalian cell. Wherein associated means co-purified with a recombinant protein of interest produced in a mammalian cell and / or in a composition comprising a recombinant protein of interest produced in a mammalian cell. In a related aspect, the present invention relates to the use of an artificial miRNA according to the present invention for reducing hydrolase activity and / or PS degradation activity in a mammalian cell and / or in a composition comprising a recombinant protein of interest produced in a mammalian cell comprising said miRNA. Wherein reducing means compared to a control, such as compared to (the same) mammalian cell not comprising the miRNA of the present invention, preferably compared to (the same) mammalian cell transfected (or transduced) with nuclease-free water or a non-targeted siRNA (NT siRNA) or a non-targeted or empty vector variant mock transfection (or transduction). Thus, in certain embodiments, the use of an artificial miRNA according to the present invention is for reducing PS degradation activity in a composition comprising a recombinant protein of interest, wherein the recombinant protein of interest is produced in a mammalian cell comprising said artificial miRNA.

[0148] Method for producing or generating a target protein using artificial miRNA targeting two or more hydrolases and method for producing or generating a target protein using artificial miRNA targeting two or more hydrolases Methods of making or producing compositions comprising a protein of interest produced by these methods

[0149] In another aspect, the present invention relates to a method for producing a target protein, the method comprising the following steps: (a) introducing an expression vector comprising a target gene encoding a target recombinant protein into a mammalian cell; (b) introducing an artificial miRNA comprising an artificial miRNA guide strand sequence targeting two or more mammalian hydrolases acting on ester bonds into the mammalian cell, wherein the artificial miRNA is introduced into the mammalian cell before, simultaneously with, or after step (a), wherein the artificial miRNA is introduced as RNA or as a mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA; (c) culturing the mammalian cell under conditions allowing expression of the target recombinant protein; and (d) harvesting the target recombinant protein in HCCF. The artificial miRNA prevents the expression of at least two endogenous hydrolases in the mammalian cell, and / or prevents the hydrolase activity and / or polysorbate degradation activity associated with the target recombinant protein produced in the mammalian cell. Wherein association means co-purification in HCCF or with the target recombinant protein produced in the mammalian cell and / or in a composition comprising the target recombinant protein produced in the mammalian cell.

[0150] The artificial miRNA comprising the artificial miRNA guide strand sequence is combined with the mRNA (at least two mRNAs) of encoding two or more mammalian hydrolases that act on ester bonds, and suppresses their expression in mammalian cells (host cells).Two or more lytic enzymes are endogenous lytic enzymes of mammalian cells, and therefore mRNA is an endogenous mRNA expressing two or more lytic enzymes that act on ester bonds.Preferably, mammalian cells are rodent cells or human cells, and more preferably mammalian cells are rodent cells, such as CHO cells.Therefore, in certain embodiments, two or more mammalian hydrolases that act on ester bonds are rodent or human hydrolases, and preferably two or more mammalian hydrolases are rodent hydrolases such as hydrolases from CHO cells.In certain embodiments, two or more mammalian hydrolases are three or more mammalian hydrolases, four or more mammalian hydrolases or five or more mammalian hydrolases.

[0151] The term "introducing artificial miRNA" refers to any means of inserting artificial miRNA into mammalian cells, and includes transfection of RNA and transfection or transduction of a mammalian expression vector encoding a polynucleotide sequence of the artificial miRNA or comprising a polynucleotide sequence encoding the artificial miRNA. Similarly, as used herein, the term "mammalian cell comprising artificial miRNA" means a mammalian cell comprising an artificial miRNA transiently transfected as RNA, and a mammalian cell comprising a polynucleotide encoding and expressing an artificial miRNA in a mammalian cell stably inserted into the genome or transiently transfected.

[0152] The HCCF of step (d) has reduced hydrolase activity and / or reduced PS degradation activity, preferably compared with the HCCF of the target recombinant protein produced in mammalian cells that do not contain (or encode) artificial miRNA (miRNA), the HCCF has reduced hydrolase activity and / or reduced PS degradation activity. Methods for determining reduced hydrolase activity and / or reduced PS degradation activity are known in the art, and include methods described herein. Compared with the HCCF of the target recombinant protein produced in mammalian cells that do not contain (or encode) artificial miRNA (miRNA), the HCCF of step (d) may also have reduced levels of two or more mRNAs encoding two or more hydrolases that act on ester bonds. The mRNA level can be easily measured using reverse transcription polymerase chain reaction (RT-PCR), preferably using quantitative PCR (qPCR). Combined RT-PCR and qPCR (quantitative RT-PCR or real-time RT-PCR) are routinely used for analysis of gene expression.

[0153] Compared to HCCF derived from mammalian cells that do not contain (or encode) artificial miRNA, the PS degradation activity in the HCCF can be reduced by 20% or more, 30% or more, 50% or more, or 70% or more. Mammalian cells (i.e., the same mammalian cells) that do not contain artificial miRNA are such as mock transfected (or transduced), such as the same mammalian cells transfected (or transduced) with nuclease-free water or non-targeted siRNA (NT siRNA) or non-targeted or empty vector variants. Mammalian cells that do not contain artificial miRNA can also be referred to as controls or control cells. It should be understood by those skilled in the art that mammalian cells are cultured under the same conditions that allow expression of the target recombinant protein and the target recombinant protein is harvested using the same method. Reduced PS degradation activity can refer to a reduction in PS concentration (PS20 and / or PS80) on a certain day, such as the 7th day or the 14th day at room temperature, and can be provided in absolute or relative values, including % control, or refers to a reduction in PS degradation rate provided in mg / ml / day and can be provided in absolute or relative values, including % control.

[0154] The method according to the present invention may further comprise the following steps: (a) purifying the target recombinant protein; and (b) formulating the purified target recombinant protein into a composition, preferably a pharmaceutical composition. A pharmaceutical composition refers to a composition comprising the target recombinant protein and at least one pharmaceutically acceptable excipient.

[0155] In certain embodiments, the pharmaceutical composition comprises PS, such as PS20 or PS80. Therefore, PS20 or PS80 and optionally another excipient can be added. In certain embodiments, the composition comprises PS, and when the composition is stored at about 2°C to about 8°C, preferably at about 5°C for at least six months, less than about 10%, less than about 5%, or less than about 2%, or less than about 1% of PS degradation, and / or when stored at about 2°C to about 8°C, preferably at about 5°C for at least 12 months, less than about 15%, less than about 10%, less than about 5%, or less than about 2% of PS degradation. 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 of PS20 or PS80 is added, more preferably about 0.4% (w / v) or more of PS20 or PS80, more preferably about 0.4% (w / v) of PS20.

[0156] The composition comprising the purified recombinant protein of interest of step (e) or the composition of step (f) has reduced hydrolase activity and / or reduced PS degradation activity, preferably compared to a composition comprising a purified recombinant protein of interest produced in a mammalian cell that does not contain or encodes an artificial miRNA. The hydrolase activity or reduced PS degradation activity can be attributed to contaminating HCPs co-purified with the protein of interest, such as hydrolases acting on esterase bonds. The composition comprising the purified recombinant protein of interest may also have reduced immunogenic activity. The artificial miRNA according to the present invention prevents the expression of at least two endogenous hydrolases in mammalian cells, and / or prevents the hydrolase activity and / or polysorbate degradation activity associated with the recombinant protein of interest produced in mammalian cells. Wherein associated means co-purified with the recombinant protein of interest produced in a mammalian cell and / or in a composition comprising the recombinant protein of interest produced in a mammalian cell. The composition in this context may be any composition comprising the recombinant protein of interest (such as an antibody), in particular HCCF, a composition comprising the purified recombinant protein of interest of step (d) or a composition of step (e), in particular a pharmaceutical composition comprising a pharmaceutically acceptable excipient.

[0157] Some lipases such as PLBL2 are reported to be immunogenic in patients. Therefore, reducing the expression of two or more hydrolases (such as lipases and / or carboxylesterases) acting on ester bonds in mammalian host cells can also lead to reduced immunogenic activity of purified recombinant proteins of interest or compositions thereof produced in mammalian host cells. Therefore, the artificial miRNA and methods of the present invention can further improve patient safety.

[0158] In certain embodiments, the expression vector comprising the target gene encoding the target recombinant protein is introduced by transfection or transduction, preferably by transfection. Preferably, the target gene is stably introduced by transfection or transduction. More preferably, the expression vector comprising the target gene encoding the target recombinant protein is stably transfected into mammalian cells.

[0159] The artificial miRNA is introduced as RNA or as a mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA. The artificial miRNA is introduced as RNA by transient transfection, or by transfection or transduction, preferably by stable transfection or transduction, more preferably by stable transfection as an expression vector comprising a polynucleotide sequence encoding the artificial miRNA. In certain embodiments, the artificial miRNA is introduced into mammalian cells by stable transfection or transduction of an expression vector comprising a polynucleotide sequence encoding the artificial miRNA, wherein the expression vector comprising the polynucleotide sequence encoding the artificial miRNA also comprises a target gene encoding a target recombinant protein. Stable integration (i.e., transfection or transduction) includes semi-targeted, targeted and random integration. The expression vector and / or artificial miRNA introduced according to the method of the present invention is preferably a mammalian expression vector and / or an artificial miRNA according to the present invention, or the expression vector comprises a polynucleotide sequence encoding the artificial miRNA according to the present invention. Where applicable, the disclosures about artificial miRNA, mammalian expression vectors and / or mammalian cells according to the present invention are similarly applicable to the methods of the present invention.

[0160] More specifically, in certain embodiments, the two or more mammalian hydrolases are three or more mammalian hydrolases, four or more mammalian hydrolases, five or more mammalian hydrolases, six or more mammalian hydrolases, seven or more mammalian hydrolases, or eight or more mammalian hydrolases.

[0161] The artificial miRNA guide strand sequence has incomplete complementarity with at least one of the at least two mRNAs encoding two or more mammalian hydrolases, preferably with more than one target sequence in the at least two mRNAs encoding two or more mammalian hydrolases. Therefore, the artificial miRNA guide strand sequence has complete complementarity with the target sequence of one mRNA of at least two mRNAs encoding two or more mammalian hydrolases, and has incomplete complementarity with one or more target sequences of the remaining mRNAs of at least two mRNAs encoding two or more (targeted) mammalian hydrolases. In certain embodiments, the artificial miRNA guide strand has incomplete complementarity with one or more target sequences of all at least two mRNAs encoding two or more (targeted) mammalian hydrolases, or the artificial miRNA guide strand has incomplete complementarity with one or more target sequences of all or at least two mRNAs encoding two or more (targeted) mammalian hydrolases except one. Preferably, the artificial microRNA (miRNA) according to the present invention comprises an artificial miRNA guide strand sequence targeting three, four, five, six, seven, eight, nine or more mammalian hydrolases acting on ester bonds. In certain embodiments, the artificial miRNA guide strand has incomplete complementarity with one or more target sequences of mRNAs encoding all or all but one of three, four, five, six, seven, eight, nine or more (targeted) mammalian hydrolases, preferably CHO hydrolases. It will be appreciated by those skilled in the art that the artificial miRNA guide strand sequence may have complete complementarity with the target sequence of one mRNA of the mRNAs encoding three, four, five, six, seven, eight, nine or more mammalian hydrolases, preferably CHO hydrolases.

[0162] The target sequence may be located anywhere in the mRNA sequence encoding at least two mRNAs encoding two or more mammalian hydrolases, preferably in the 3' untranslated region or coding region of at least two mRNAs encoding two or more mammalian hydrolases, more preferably in the coding region of at least two mRNAs encoding two or more mammalian hydrolases, and even more preferably in a microconserved region in the coding region of at least two mRNAs encoding two or more mammalian hydrolases. Typically, the target sequence of all at least two mRNAs encoding two or more mammalian hydrolases is in its 3' untranslated region or coding region, preferably in the coding region of at least two mRNAs encoding two or more mammalian hydrolases, and even more preferably in a microconserved region in the coding region of at least two mRNAs encoding two or more mammalian hydrolases.

[0163] In certain embodiments, the target sequence is located in a conserved region of the amino acid sequence of two or more mammalian hydrolases. In certain preferred embodiments, the target sequence is located in a conserved region of at least two mRNAs encoding two or more mammalian hydrolases, preferably in a conserved region of the coding region of at least two mRNAs encoding two or more mammalian hydrolases. As used herein, the term "conserved region" refers to identical or similar nucleotide sequences in at least two mRNAs encoding two or more mammalian hydrolases. Wherein similar sequences mean a pairwise identity percentage of more than 55%. Pairwise identity refers to the percentage of identical paired residues in the alignment, including gaps relative to non-gap residues, but excluding gaps relative to gap residues. A microconserved region is an identical or similar sequence of a very small segment of at least 21 nucleotides, preferably with at least 55% sequence identity.

[0164] The artificial miRNA guide strand sequence of the artificial miRNA used in the method of the present invention typically has 19 to 25 nucleotides, preferably 20 to 24 nucleotides, more preferably 21 to 23 nucleotides, and even more preferably 22 to 23. Binding to one or more target sequences of mRNA requires binding of the 5' seed region and optionally the 3' complementary region (rather than the entire guide strand sequence). Therefore, preferably the artificial miRNA guide strand sequence comprises at least a 5' seed region and a 3' complementary region. In certain embodiments, the artificial miRNA guide strand sequence does not have complete complementarity with at least one of the at least two mRNAs encoding two or more mammalian hydrolases, and the miRNA guide strand has a 5' seed region and a 3' complementary region that undergoes base pairing with the one or more target sequences, preferably a 5' seed region starting at nucleotide 2 (nt 2) at the 5' end of the artificial miRNA guide strand sequence. In general, the artificial miRNA guide strand sequence can have at least 55%, preferably at least 60%, and more preferably at least 65% sequence complementarity with one or more target sequences of at least two mRNAs encoding two or more mammalian hydrolases. In certain embodiments, the miRNA guide strand sequence has at least 13, preferably 14, and more preferably 15 Watson-Crick base pairings (continuous and non-continuous) with one or more target sequences of at least two mRNAs encoding two or more mammalian hydrolases. Thus, the 22 to 23 nucleotide artificial miRNA guide strand sequence comprises 8-10, preferably 8-9 G or C complementary to (i.e., undergoing base pairing with) one or more target sequences. In certain preferred embodiments, the first nucleotide at the 5' end of the artificial miRNA guide strand sequence (i.e., the 5' nucleotide of the 5' seed region) comprises a nucleotide (comprising a nucleobase) selected from adenine (A) and uracil (U), preferably the nucleotide is U.

[0165] The two or more mammalian hydrolases acting on ester bonds may be any mammalian hydrolases, in particular two or more mammalian lipases or carboxylesterases and / or two or more PS degrading mammalian hydrolases.

[0166] The inventors have identified two types of artificial miRNA guide strand sequences that effectively reduce the expression and / or activity of hydrolases. The first sequence type comprises a strong 5' seed region, where strong means high GC content, high complementarity (≥62.5% sequence complementarity, preferably ≥75% sequence complementarity, or no more than 3 mismatches in 8 nucleotides, preferably no more than 2 mismatches in 8 nucleotides); and a weak or normal 3' complementary region (if present) that can compensate for mismatches in the seed region. The second sequence comprises a weaker 5' seed region, which is attributed to a lower GC content or a shorter complementary region, particularly such as having continuous mismatches at the ends of the region; and a strong 3' complementary region that compensates for weaker binding and / or mismatches in the 5' seed region.

[0167] In certain embodiments of the methods according to the invention, the miRNA guide strand sequence has incomplete complementarity with one or more target sequences of at least one of the at least two mRNAs encoding two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) a sequence of 8 nucleotides (nt 2-9) that is complementary to the one or more target sequences and has 0, 1, 2 or 3 mismatches with the one or more target sequences, and (ii) a GC content of 55-70%, wherein all or all but one or all but two of the Gs or Cs are complementary to the one or more target sequences; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides (i.e., nt-6 to nt-1) that is complementary to the one or more target sequences and has 1, 2 or 3 mismatches with the one or more target sequences, and (ii) a GC content of ≤50% in the last 6 nucleotides, wherein 1 or 2 complementary nucleotides are G or C. Examples of suitable artificial miRNA guide strand sequences are the nucleotide sequences of SEQ ID NO: 2 (LP 6mer), SEQ ID NO: 4 (LH 9mer) or SEQ ID NO: 5 (non-classical LH). Preferably, the artificial miRNA guide strand sequence used in the method according to the present invention is the nucleotide sequence of SEQ ID NO: 2 (LP6mer) or SEQ ID NO: 5 (non-classical LH).

[0168] In certain alternative embodiments of the methods according to the invention, the miRNA guide strand sequence has incomplete complementarity with one or more target sequences of at least one of the at least two mRNAs encoding two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) a sequence of 8 nucleotides (nt 2-9), wherein at least the first 4 consecutive nucleotides (at least nt 2-5) are complementary to the one or more target sequences, wherein the sequence of 8 nucleotides comprises at least one mismatch with the one or more target sequences, and (ii) a GC content of 55-75%, wherein at least 4 consecutive nucleotides complementary to the one or more target sequences are G or C; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides that are complementary to the one or more target sequences and have 0 or 1 mismatches with the one or more target sequences, and (ii) a GC content of 30-50% in the last 6 nucleotides, preferably wherein at least 2 complementary nucleotides are G or C. An example of a suitable artificial miRNA guide strand sequence is the nucleotide sequence of SEQ ID NO: 1 (LP 8mer) or SEQ ID NO: 2 (LP 6mer).

[0169] The artificial miRNA guide strand sequence according to the method of the present invention is designed from scratch, which means that it is not naturally present (or not identified) in mammalian cells or organisms. Therefore, the sequence is generated by rational design based on sequence alignment of the target sequence at the mRNA and / or protein level (preferably at the mRNA level). The object of the present invention is to provide the following proof of concept: artificial miRNAs comprising artificial miRNA guide strand sequences can be used in cell line engineering methods for simultaneously targeting multiple hydrolases acting on ester bonds, such as multiple lipases, in mammalian host cells such as CHO cells.

[0170] The artificial miRNA according to the method of the present invention can be pri-miRNA, pre-miRNA, double-stranded mature miRNA (comprising a follower strand and a guide strand) or a single-stranded miRNA guide strand. In certain embodiments, the artificial miRNA is a single-stranded or double-stranded miRNA mimic (miR-Mimic). MiRNA mimics are chemically synthesized double-stranded RNAs that simulate mature miRNAs after transfection into cells and can be provided as double-stranded miRNA mimics or as single-stranded miRNAs, the latter requiring an additional annealing step with the follower strand before transfection. Customized miRNA mimics can be obtained by proprietary chemical modifications to improve their performance and stability and to prevent the binding activity of follower strands obtained on demand from different suppliers (such as from Horizon, Qiagen or Thermo Fisher).

[0171] The two or more mammalian hydrolases acting on ester bonds targeted by the artificial miRNA according to the present invention are endogenous hydrolases of mammalian cells, preferably the mammalian cells are human or rodent cells, more preferably rodent cells, such as CHO cells.

[0172] According to the method of the present invention, a target protein, such as an antibody, is produced in a mammalian cell (e.g., a CHO cell) in a cell culture. After expression, the recombinant protein is harvested and further purified. The target recombinant protein, such as an antibody, can be recovered from the culture medium as a secretory protein in the HCCF, or recovered from a cell lysate (i.e., a fluid containing cell contents cracked by any means (including but not limited to enzymatic, chemical, osmotic, mechanical and / or physical destruction of the cell membrane and optionally the cell wall)), and purified using the techniques described herein. According to the present invention, the method includes harvesting the target recombinant protein in the HCCF. Preferably, after separating the cells such as by filtering and / or centrifugation, the target recombinant protein (e.g., antibody) is recovered (i.e., purified) from the HCCF. Therefore, in certain embodiments, harvesting includes centrifugation and / or filtration to produce 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. In addition, filtration, such as membrane and / or deep filtration, can also be applied.

[0173] The present invention also encompasses a method for producing a cell line that stably expresses the artificial miRNA of the present invention, the method comprising the following steps: (a) stably introducing a mammalian expression vector comprising a polynucleotide sequence encoding an artificial miRNA into mammalian cells, wherein the artificial miRNA comprises an artificial miRNA guide strand sequence that targets two or more mammalian hydrolases for ester bonds, and (b) selecting mammalian cell clones that stably express the artificial miRNA, and optionally using the cells for cell line generation, and comprising (c) introducing an expression vector comprising a target gene encoding a target recombinant protein into mammalian cells that stably express the artificial miRNA.

[0174] In another aspect, the present invention relates to a composition comprising a recombinant protein of interest, wherein the recombinant protein of interest can be obtained by a method according to the present invention. Preferably, the composition further comprises PS (PS20 or PS80), preferably at a concentration of 0.1 g / L or more. In certain embodiments, the composition comprises PS, and when the composition is stored at about 2°C to about 8°C, preferably at about 5°C for at least six months, less than about 10% of the PS is degraded.

[0175] The composition comprising the target recombinant protein according to the present invention has reduced hydrolase activity and / or reduced PS degradation activity. Preferably, the composition comprising the target recombinant protein has reduced hydrolase activity and / or reduced PS degradation activity compared to a composition comprising the target recombinant protein produced in a mammalian cell that does not contain (or encodes) an artificial miRNA.

[0176] In view of the above, it should be understood that the present invention also covers the following items:

[0177] Item 1 provides an artificial micro RNA (miRNA), which comprises an artificial miRNA guide strand sequence, wherein the artificial miRNA guide strand sequence targets two or more mammalian hydrolases acting on ester bonds.

[0178] Item 2 further specifies the artificial miRNA according to Item 1, wherein the two or more hydrolases are endogenous hydrolases of mammalian cells, preferably human or rodent cells, more preferably rodent cells, and even more preferably CHO cells.

[0179] Item 3 further specifies the artificial miRNA according to Item 1 or 2, wherein the artificial miRNA guide strand sequence has complete or incomplete complementarity with the target sequences of at least two mRNAs encoding the two or more mammalian hydrolases.

[0180] Item 4 further specifies the artificial miRNA according to Item 3, wherein the artificial miRNA guide strand sequence has incomplete complementarity with at least one of the at least two mRNAs encoding the two or more mammalian hydrolases, preferably with the one or more target sequences encoding more than one of the at least two mRNAs encoding the two or more mammalian hydrolases. Preferably, the artificial miRNA guide strand sequence has incomplete complementarity with all or all but one of the target sequences of the at least two mRNAs encoding the two or more mammalian hydrolases.

[0181] Item 5 further specifies the artificial miRNA according to Item 3 or 4, wherein the target sequence is located in the 3'UTR or coding region of the at least two mRNAs encoding the two or more mammalian hydrolases, preferably in the coding region of the at least two mRNAs encoding the two or more mammalian hydrolases, more preferably in a micro-conserved region in the coding region of the at least two mRNAs encoding the two or more mammalian hydrolases.

[0182] Item 6 further specifies the artificial miRNA according to Item 5, wherein the target sequence is located in a conserved region of the at least two mRNAs encoding the two or more mammalian hydrolases, preferably in a conserved region of the coding regions of the at least two mRNAs encoding the two or more mammalian hydrolases.

[0183] Item 7 further specifies the artificial miRNA according to any one of items 1 to 6, wherein the artificial miRNA targeting two or more mammalian hydrolases reduces the activity of the two or more hydrolases in mammalian cells, preferably human or rodent cells, more preferably rodent cells, and even more preferably CHO cells.

[0184] Item 8 further specifies the artificial miRNA according to any one of the preceding items, wherein the artificial miRNA guide strand sequence has 19 to 25 nucleotides, preferably 20 to 24 nucleotides, more preferably 21 to 23 nucleotides.

[0185] Item 9 further specifies the artificial miRNA according to any of the preceding items, wherein the artificial miRNA guide strand sequence independently hybridizes with the target sequences of the at least two mRNAs encoding the two or more mammalian hydrolases without forming a bulge loop and / or forming at most two bulge loops, preferably without forming a bulge loop.

[0186] Item 10 further specifies an artificial miRNA according to any one of Items 1 to 9, wherein the artificial miRNA guide strand sequence has at least 55% sequence complementarity with the one or more target sequences of the at least two mRNAs encoding the two or more mammalian hydrolases; and / or wherein the miRNA guide strand sequence has at least 13 Watson-Crick base pairings with the one or more target sequences of the at least two mRNAs encoding the two or more mammalian hydrolases.

[0187] Item 11 further specifies the artificial miRNA according to any one of items 1 to 10, wherein one of the two or more mammalian hydrolases is lipoprotein lipase (LPL), and preferably wherein one of the two or more mammalian hydrolases is LPL, and the one or more target sequences in the mRNA encoding LPL are within nucleotides 560-590, 700-745, 930-960 or 1330-1360 of SEQ ID NO: 6 or 7, preferably within nucleotides 560-590, 705-745, or 930-960 of SEQ ID NO: 6 or 7.

[0188] Item 12 further specifies the artificial miRNA according to any one of items 1 to 11, wherein the two or more mammalian hydrolases are selected from lipoprotein lipase (LPL), phospholipase B-like protein 2 (PLBL2), phospholipase A1 (Pla1a), inactive pancreatic lipase-related protein 1 (PNLIPRP1), inactive pancreatic lipase-related protein 2 (PNLIPRP2), pancreatic triglyceride lipase precursor (PNLIP), lipase member I precursor (LipI), lipase member H precursor (LipH), lipase member G precursor (LipG), lipase member C precursor (LipC) and their subtypes.

[0189] Item 13 further specifies the artificial miRNA according to any one of items 1 to 12, wherein the artificial miRNA guide strand sequence has a nucleotide sequence of SEQ ID NO: 1 (LP 8mer), SEQ ID NO: 2 (LP 6mer), SEQ ID NO: 3 (LH8mer), SEQ ID NO: 4 (LH 9mer) or SEQ ID NO: 5 (non-classical LH).

[0190] Item 14 further specifies an artificial miRNA according to any one of the preceding items, wherein the artificial miRNA guide strand sequence has incomplete complementarity with the one or more target sequences of at least one of the at least two mRNAs encoding the two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) 8 nucleotides (nt 2-9), which sequence is complementary to the one or more target sequences and has 0, 1, 2 or 3 mismatches, preferably 0, 1 or 2 mismatches, more preferably 0 or 1 mismatch, and (ii) a GC content of 55-70%, wherein all or all except one or all except two Gs or Cs are complementary to the one or more target sequences; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides, which sequence is complementary to the one or more target sequences and has 1, 2 or 3 mismatches with the one or more target sequences, preferably wherein the mismatches are not consecutive mismatches, and (ii) a GC content of ≤50% in the last 6 nucleotides, wherein 1 or 2 of the complementary nucleotides are G or C.

[0191] Item 15 further specifies the artificial miRNA according to Item 14, wherein (a) the 5' seed region does not contain mismatches within nucleotides 2-3, preferably within nucleotides 2-4, more preferably within nucleotides 2-5, and even more preferably within nucleotides 2-7, and / or (b) the 5' seed sequence contains at least 2 times of 3-4 consecutive base pairings with the one or more target sequences, and / or (c) wherein the mismatches in the 5' seed region are not consecutive mismatches.

[0192] Item 16 further specifies the artificial miRNA according to item 14 or 15, wherein the artificial miRNA guide strand sequence has the nucleotide sequence of SEQ ID NO: 2 (LP 6mer), SEQ ID NO: 4 (LH 9mer) or SEQ ID NO: 5 (non-classical LH).

[0193] Item 17 further specifies an artificial miRNA according to any one of items 1 to 13, wherein the artificial miRNA guide strand sequence has incomplete complementarity with the one or more target sequences of at least one of the at least two mRNAs encoding the two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) a sequence of 8 nucleotides (nt 2-9), wherein at least the first 4 consecutive nucleotides (at least nt 2-5), preferably at least the first 6 consecutive nucleotides (at least nt 2-7) complementary to said one or more target sequences, wherein said 8 nucleotide sequence comprises at least one mismatch with said one or more target sequences, and (ii) a GC content of 55-75%, wherein at least 4 of said consecutive nucleotides complementary to said one or more target sequences are G or C; and wherein said 3' complementary region comprises (i) a sequence of the last 6 nucleotides, which is complementary to said one or more target sequences and has 0 or 1 mismatch with said one or more target sequences, preferably wherein the last nucleotide is complementary to said one or more target sequences and / or the second to last nucleotide is a mismatch, and (ii) a GC content of 30-50% in said last 6 nucleotides, preferably wherein at least 2 of said complementary nucleotides are G or C.

[0194] Item 18 further specifies the artificial miRNA according to Item 17, wherein the artificial miRNA guide strand sequence has a nucleotide sequence of SEQ ID NO: 1 (LP 8mer) or SEQ ID NO: 2 (LP 6mer).

[0195] Item 19 further specifies the artificial miRNA according to any one of items 1 to 18, wherein the first nucleotide at the 5' end of the artificial miRNA guide strand sequence comprises a nucleotide selected from A and U, preferably the nucleotide is U.

[0196] Item 20 further specifies the artificial miRNA according to any one of items 1 to 19, wherein the artificial miRNA is a pri-miRNA, a pre-miRNA, a double-stranded mature miRNA (comprising a follower strand and a guide strand) or a single-stranded miRNA guide strand.

[0197] Item 20 further specifies the artificial miRNA according to any one of items 1 to 20, wherein the artificial miRNA guide strand sequence is designed de novo.

[0198] Item 22 further specifies the artificial miRNA according to any one of items 1 to 21, wherein the artificial miRNA is a single-stranded or double-stranded miRNA mimic.

[0199] Item 23 provides a mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA according to any one of Items 1 to 21.

[0200] Item 24 further specifies that the mammalian expression vector according to Item 23 further comprises at least one target gene encoding a target recombinant protein.

[0201] Item 25 further specifies a mammalian expression vector according to item 23 or 24, wherein the mammalian expression vector comprises a polynucleotide sequence encoding two or more artificial miRNAs according to any one of items 1 to 21, wherein the artificial miRNAs are the same or different.

[0202] Item 26 provides a mammalian cell, preferably a CHO cell, comprising the artificial miRNA according to any one of items 1 to 22, or a polynucleotide encoding the artificial miRNA according to any one of items 1 to 22, or a mammalian expression vector according to any one of items 23 to 25.

[0203] Item 27 further specifies the mammalian cell according to Item 26, wherein the mammalian cell is a human cell or a rodent cell, preferably a CHO cell.

[0204] Item 28 further specifies the mammalian cell according to item 26 or 27, wherein the mammalian cell also expresses a recombinant protein of interest, preferably a therapeutic protein, more preferably a therapeutic antibody.

[0205] Item 29 further specifies the mammalian cell according to Item 28, wherein the target recombinant protein is stably expressed.

[0206] Item 30 further specifies a mammalian cell according to any one of items 26 to 29, wherein (a) the mammalian cell has reduced expression of two or more mammalian hydrolases compared to a mammalian cell that does not contain the artificial miRNA according to any one of items 1 to 22; (b) the mammalian cell and / or the recombinant protein of interest produced and isolated from the mammalian cell has reduced hydrolase activity compared to a mammalian cell that does not contain the artificial miRNA according to any one of items 1 to 22 and / or the recombinant protein of interest produced and isolated from the mammalian cell; (c) the mammalian cell and / or the recombinant protein of interest produced and isolated from the mammalian cell has reduced hydrolase activity compared to a mammalian cell that does not contain the artificial miRNA according to any one of items 1 to 22 and / or the recombinant protein of interest produced and isolated from the mammalian cell. (a) the mammalian cell and / or the recombinant protein of interest produced and isolated from the mammalian cell has reduced PS degradation activity of two or more mammalian hydrolases compared to a recombinant protein of interest produced and isolated from the mammalian cell; and / or (d) the mammalian cell and / or the recombinant protein of interest produced and isolated from the mammalian cell has reduced PS degradation activity compared to a mammalian cell and / or a recombinant protein of interest produced and isolated from the mammalian cell that does not contain the artificial miRNA according to any one of items 1 to 22; and / or (e) the artificial miRNA prevents the expression of at least two endogenous hydrolases in the mammalian cell, and / or prevents the hydrolase activity and / or polysorbate degradation activity associated with the recombinant protein of interest produced in the mammalian cell.

[0207] Item 31 further specifies a mammalian cell according to any one of items 26 to 30, wherein the mammalian cell is transfected or transduced with a mammalian expression vector according to any one of items 23 or 25, preferably stably transfected or transduced, including semi-targeted, targeted and random integration.

[0208] Item 32 further specifies a mammalian cell according to any one of items 26 to 30, wherein the mammalian cell is transfected with the artificial miRNA according to any one of items 1 to 22 or a mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA according to any one of items 1 to 21.

[0209] Item 33 provides use of the mammalian cell according to any one of items 26 to 32 for producing a recombinant protein of interest.

[0210] Item 34 provides the use of the artificial miRNA according to any one of items 1 to 22 in a mammalian cell for preventing the expression of at least two endogenous hydrolases in the mammalian cell and / or preventing the hydrolase activity and / or polysorbate degradation activity associated with a recombinant protein of interest produced in the mammalian cell.

[0211] Item 35 provides use of the artificial miRNA according to any one of items 1 to 22 for reducing PS degradation activity in a composition comprising a target recombinant protein, wherein the target recombinant protein is produced in a mammalian cell comprising the artificial miRNA.

[0212] Item 36 provides a method for producing a target protein, comprising the following steps: (a) introducing an expression vector containing a target gene encoding a target recombinant protein into mammalian cells; (b) introducing an artificial miRNA containing an artificial miRNA guide chain sequence targeting two or more mammalian hydrolases acting on ester bonds into the mammalian cells, wherein the artificial miRNA is introduced into the mammalian cells before, simultaneously with or after step (a), wherein the artificial miRNA is introduced as RNA or as a mammalian expression vector containing a polynucleotide sequence encoding the artificial miRNA; (c) culturing the mammalian cells under conditions allowing expression of the target recombinant protein; and (d) harvesting the target recombinant protein in HCCF.

[0213] Item 37 further specifies the method according to Item 36, wherein the HCCF has reduced hydrolase activity and / or reduced PS degradation activity, preferably compared to HCCF containing the target recombinant protein produced in mammalian cells that do not contain or encode the artificial miRNA (miRNA), the HCCF has reduced hydrolase activity and / or reduced PS degradation activity.

[0214] Item 38 further specifies the method according to item 36 or 37, wherein the PS degradation activity in the HCCF is reduced by 30% or more, 50% or more, or 70% or more compared to HCCF derived from mammalian cells that does not contain or encode the artificial miRNA.

[0215] Item 39 further specifies a method according to any one of items 36 to 38, wherein the method further comprises the following steps: (e) purifying the target recombinant protein; and optionally (f) formulating the purified target recombinant protein into a composition, preferably a pharmaceutical composition.

[0216] Item 40 further specifies the method according to item 39, wherein the composition comprises PS (PS20 or PS80), and wherein less than about 10% of the PS degrades when the composition is stored at about 2°C to about 8°C for at least six months.

[0217] Item 41 further specifies the method according to Item 39 or 40, wherein the composition comprising the purified recombinant protein of interest according to step (e) or the composition of step (f) has reduced hydrolase activity and / or reduced PS degradation activity, preferably compared to a composition comprising the purified recombinant protein of interest produced in mammalian cells that does not contain or encodes the artificial miRNA, the composition comprising the purified recombinant protein of interest has reduced hydrolase activity and / or reduced PS degradation activity.

[0218] Item 42 further specifies the method according to any one of items 36 to 41, wherein the expression vector comprising the gene of interest encoding the recombinant protein of interest is introduced by transfection or transduction.

[0219] Item 43 further specifies the method according to any one of items 36 to 42, wherein the artificial miRNA is introduced as RNA by transient transfection, or introduced as an expression vector comprising a polynucleotide sequence encoding the artificial miRNA by transfection or transduction.

[0220] Item 44 further specifies the method according to Item 43, wherein the artificial miRNA is introduced into the mammalian cells by stable transfection or transduction of an expression vector comprising a polynucleotide sequence encoding the artificial miRNA, optionally wherein the expression vector comprising the polynucleotide sequence encoding the artificial miRNA also comprises the target gene encoding the target recombinant protein.

[0221] Item 45 further specifies the method according to any one of items 36 to 44, wherein the expression vector comprising the polynucleotide sequence encoding the artificial miRNA is a mammalian expression vector according to any one of items 23 to 25.

[0222] Item 46 further specifies the method according to any one of items 36 to 45, wherein the artificial miRNA is the artificial miRNA according to any one of items 1 to 22.

[0223] Item 47 further specifies a method according to any one of items 36 to 46, wherein the two or more mammalian hydrolases are endogenous hydrolases of the mammalian cells, preferably the mammalian cells are human or rodent cells, more preferably CHO cells.

[0224] Item 48 provides a composition comprising a target recombinant protein and preferably PS at a concentration of 0.1 g / L or greater, wherein the target recombinant protein can be obtained by the method according to any one of items 35 to 46.

[0225] Item 49 further specifies the composition according to item 48, wherein the composition comprises PS (PS20 or PS80), and wherein less than about 10% of the PS degrades when the composition is stored at about 2°C to about 8°C for at least six months.

[0226] Item 50 further specifies the composition according to Item 48 or 49, wherein the composition comprising the target recombinant protein has reduced hydrolase activity and / or reduced PS degradation activity, preferably compared to a composition comprising the target recombinant protein produced in mammalian cells without introducing or comprising the artificial miRNA, the composition comprising the target recombinant protein has reduced hydrolase activity and / or reduced PS degradation activity.

[0227] Example

[0228] Materials and methods

[0229] Cell lines and culture

[0230] The experiments were performed using a CHO-K1 cell line with biallelic GS knockout (CHO-K1 GS) as well as CHO-K1 GS cells expressing different recombinant products.

[0231] CHO cells are suitable for growth in chemically defined and animal component-free media in suspension culture. For standard culture, cells were cultured in 60 ml of medium using 250 ml shake flasks (Corning, Corning, USA) and incubated in a Multitron incubator (Inforse, Bottmingen, Switzerland) at 37°C, 5% CO 2 and 125 rpm (50 mm orbit). Use the corresponding inoculation density (0.4 × 10 6 / 0.2×10 6 CHO cells were passaged at a rhythm of 2 / 2 / 3 days (10 cells / ml (2d / 3d)).

[0232] Measurement of live cell concentration and viability

[0233] To measure the viable cell concentration (VCC) and viability of cells cultured in shake flasks, InnovatisCedexTM (Roche Diagnostics, Mannheim, Germany) was used. For each measurement, 1 ml of the cell suspension was pipetted into a designated sample container and the fully automated measurement was started. Image-based technology uses the trypan blue (Life Technologies, Carlsbad, CA USA) exclusion method for cell counting. Dead cells with damaged cell membranes will be stained with trypan blue, while live cells with intact cell membranes will exclude the dye (Strober 2001).

[0234] To measure the VCC and viability of cells cultured in 96 half-deep well (DW) plates (EnzyScreen BV, Heemstede, Netherlands) or centrifuge tubes (TPP, Trasadingen, Switzerland), a NovoCyte flow cytometer (Acea, San Diego, USA) was used. For each measurement, 25 μl of cell suspension was mixed with 75 μl of propidium iodide solution (PI; c = 6, 67 μg / ml) (Sigma Aldrich, Steinheim, Germany) for viability staining. Subsequently, 1×10 cells were analyzed by plotting forward scatter (488 nm laser, 488 / 10 nm filter) and side scatter (488 nm laser, 488 / 10 nm filter). 4 events. In addition, the PI content of the cells was measured by an additional laser-filter combination (488 nm laser, 660 / 20 nm filter). PI is embedded in the double-stranded DNA of dying cells and is therefore used as an indicator of apoptosis (Lecoeur 2002). After batch cultivation of amiRNA-transfected CHO production cell lines, the VCC was used to calculate the integral viable cell concentration (IVC).

[0235] Posttranscriptional gene regulation via double-stranded RNA

[0236] Double-stranded RNA (dsRNA) molecules (such as amiRNA mimics and siRNA, applied as positive and negative controls) are used for post-transcriptional gene regulation in CHO cell lines. Therefore, dsRNA was transfected using the NeonTM transfection system and the associated 10 μl NeonTM kit (ThermoFisher, Waltham, USA). One day before transfection, CHO cells were cultured at 0.6×10 6 Each time the cells were transfected, 1.5×10 5The cells were centrifuged at 200 x g for 7 minutes and washed with 1x phosphate buffered saline (PBS) (Life Technologies, Waltham, USA). The cells were resuspended in resuspension buffer R (10 μl NeonTM kit) and mixed with 9 pmol dsRNA to obtain a total of 10 μl transfection mixture. In addition, 3 ml of electroporation buffer E (10 μl NeonTM kit) was pipetted into the electroporation tube of the transfection station. The electroporation buffer was replaced before the new dsRNA electroporation, and the electroporation tube was replaced after 10 transfections. The cells were electroporated with the following transfection settings: 1700 V, 20 milliseconds, 1 pulse. The cells were transferred to 290 μl pre-warmed culture medium in a square 96 half-DW microplate and incubated at 37 ° C and 5% CO 2 After static incubation, the microplate was placed in a shaking incubator at 37°C and 5% CO 2 Incubate at 37°C, 5% CO, 90% humidity and 225 rpm (50 mm orbit) until harvest on day 3 (72 hours). For transfections on a 1 ml scale, 3 times the number of cells and dsRNA were used per transfection. 2 The dsRNAs used in this work are listed in Table 3.

[0237] Table 3: Transfected miRNA / siRNA and corresponding sequence, target(s) and supplier.

[0238]

[0239] Cell lysis by freeze-thaw

[0240] For cell lysis, 1.0 × 10 6 The cells were precipitated at 300 x g for 7 minutes and resuspended in 11.2 μl PBS supplemented with cOmpleteTM (Sigma Aldrich, Steinheim, Germany). The cell suspension was frozen in liquid nitrogen for 2 minutes, thawed in a 36°C water bath, and vortexed thoroughly. The freeze-thaw cycle was repeated 3 times. For harvesting, the lysate was centrifuged in a desktop centrifuge at 4°C and full speed for 10 minutes. The lysate was transferred to a new reaction tube and stored at -70°C. The total protein content of the cell lysate sample was quantified via the Bradford protein quantification method.

[0241] LPL sandwich ELISA

[0242] Lipase enzyme-linked immunosorbent assay (ELISA) was performed using a commercially available ELISA kit (Cloud-Clone Corp) to quantify LPL. A standard dilution series ranging from 1.56ng / ml to 100ng / ml was prepared by diluting a 100ng / ml LPL stock solution with a standard diluent. The anti-LPL antibody pre-coated wells were filled with 100μl of standard or undiluted HCCF, sealed with a plate sealer and incubated at 37°C for 1 hour. After incubation, the liquid was removed from each well and 100μl of detection reagent A (biotin-conjugated secondary antibody) was added to each well. The sealed plates were incubated at 37°C for 1 hour. The wells were washed 3x with 350μL of washing solution, and 100μl of detection reagent B (avidin-conjugated horseradish peroxidase) was added to each sample. After 5 washing steps as described above, the sealed plates were incubated at 37°C for 30 minutes. Subsequently, 90 μl of tetramethylbenzidine (TMB) substrate solution was added to each well, the plate was sealed and incubated at 37°C for 20 minutes. The enzyme-substrate reaction caused a color change in LPL-positive wells and was terminated by adding 50 μl of sulfuric acid stop solution. The color change was measured spectrophotometrically at a wavelength of 450 nm using an M 200 microplate reader, and the LPL concentration was calculated using a standard curve.

[0243] Pla1a (lipase 1a) sandwich ELISA

[0244] As an example of amiRNA downregulation of multiple lipases, Pla1a (lipase 1a) was quantified by ELISA using a commercially available kit (Elabscience). A standard dilution series ranging from 3.13ng / ml to 200ng / ml was prepared by diluting a 200ng / ml LPL stock solution with a standard diluent. The wells pre-coated with anti-Pla1a antibodies were filled with 100μl of standard or diluted cell lysate, sealed with a plate sealer and incubated at 37°C for 90 minutes. After incubation, the liquid was removed from each well and 100μl of detection reagent A (biotin-conjugated secondary antibody) was added to each well. The sealed plate was incubated at 37°C for 1 hour. The wells were washed 3x with 350μL of washing solution, and 100μl of detection reagent B (avidin-conjugated horseradish peroxidase) was added to each sample. After 5 washing steps as described above, the sealed plate was incubated at 37°C for 30 minutes. Subsequently, 90 μl of tetramethylbenzidine (TMB) substrate solution was added to each well, the plate was sealed and incubated at 37°C for 20 min. The enzyme-substrate reaction caused a color change in the Pla1a-positive wells and was terminated by adding 50 μl of sulfuric acid stop solution. The color change was measured spectrophotometrically at a wavelength of 450 nm using an M 200 microplate reader, and the Pla1a concentration was calculated using a standard curve.

[0245] Quantification of PLBL2 via biofilm interferometry

[0246] PLBL2 measurements were performed using the Octet HTX system. Purified CHO PLBL2 standards (confidential) were diluted with assay buffer (10 mM HEPES, 250 mM NaCl, 0.3% BSA, 0.05% CHAPS, pH 7.4) to prepare a standard dilution series in the range of 0.2 ng / ml-200 ng / ml. Fluorescein-anti-PLBL2 antibody (confidential) and biotin-anti-PLBL2 antibody (confidential) were diluted with assay buffer, respectively, to obtain a working concentration of 5 μg / ml. For each sample, 80 μl of PLBL2 antibody solution was pipetted into a black 384-well plate (Greiner Bio-One, Stonehouse, United Kingdom) sample plate. The sample was diluted with assay buffer, and 80 μl of the diluted sample and 80 μl of assay buffer were each added to the wells of the sample plate. Another black 384-well plate was used as a detection plate, which contained assay buffer, HRP-labeled anti-FITC antibody (c = 2 μg / ml) in assay buffer (Sigma Aldrich, Steinheim, Germany), peroxide buffer (ThermoFisher, Waltham, USA), and metal-enhanced diaminobenzidine peroxidase (DAB) (ThermoFisher, Waltham, USA) diluted 1:10 with peroxide buffer. Similar to the sample plate, 80 μl of reagent was pipetted into the well of each sample to be measured. Finally, the sample plate and the detection plate were placed in the Octet system and the fully automated measurement was started. The layer thickness was positively correlated with the PLBL2 concentration in the sample.

[0247] Fluorescent micelle assay

[0248] PS was determined by fluorescent micelle determination (FMA). PS20 and PS80 stock solutions (c = 100 mg / ml) (Croda International plc, Snaith, United Kingdom) were sterile filtered using a 0.2 μm PVDF filter (Merck Millipore, Burlington, USA). Then, PS was diluted with the corresponding culture medium, where PS20 was diluted at 1:245 and PS80 was diluted at 1:490 to obtain a PS-culture medium working solution. HCCF was mixed with the working solution at 1:50 to obtain a spiked PS20 sample (cPS20 = 0.4 mg / ml) and a spiked PS80 sample (cPS80 = 0.2 mg / ml). 200 μl was aliquoted into 5 test tubes on a cooling rack. One test tube representing sample t0 was immediately frozen at -70 ° C. The remaining test tubes were incubated in the dark at room temperature and frozen at -70 ° C after 1, 3, 7, and 14 days. Before measurement, a standard curve ranging from 0.1 to 0.6 mg / ml for PS20 and from 0.05 to 0.3 mg / ml for PS80 was generated. For PS measurement, 10 μl of sample was mixed with 240 μl of FMA reagent (150 mM NaCl, 50 mM Tris, 0.2% acetonitrile, 5 μM NPN, 0.0015% Brij-35, pH 8), followed by incubation at 35°C for 1 minute. During the incubation period, the fluorescent dye N-phenyl-1-naphthylamine (NPN) was embedded in the micelles formed by PS. 200PRO microplate reader (Tecan, Switzerland) after exciting the dye with a 350 nm laser, the fluorescence signal was measured at 420 nm and was positively correlated with the PS concentration.

[0249] Monoclonal antibody concentration determination

[0250] The product concentration of mAbs in HCCF was measured using the Octet HTX system (Forte Bio Inc, Menlo Park, USA). HCCF of IgG1 and IgG4 producing CHO cell lines were diluted with the corresponding cell culture medium and the mAb concentration was determined using a protein A sensor.

[0251] Example 1: Artificial miRNA Design: General Design Parameters

[0252] Artificial miRNA (amiRNA) is designed based on the most advanced knowledge of endogenous miRNA binding mechanism. In recent years, several key key features have been identified, which are generally used to predict miRNA target binding. After these feature analysis and evaluation, a general design parameter list suitable for amiRNA design is created. In addition, the scope of amiRNA features is defined and graded accordingly. The summary of amiRNA design criteria and parameter levels is given in Table 4.

[0253] Table 4: Summary of amiRNA design criteria and parameter levels

[0254]

[0255]

[0256] Design of artificial miRNA targeting lipase

[0257] Based on previous observations, LPL (lipase 1) and PLBL2 (lipase 2) were hypothesized to be involved in PS degradation in DS / DP and, therefore, in particle formation in formulated biotherapeutics. Therefore, genomic and transcriptomic sequence information was extracted from a database containing CHO-specific omics data. In addition, homologs were identified based on amino acid sequence homology to LPL (BlastP was performed in GenedataSelector with default settings), and sequence information for these proteins was extracted. Geneious was used to identify the genomic and transcriptomic information of the proteins. The software performed all further bioinformatics analyses, including amino acid and transcript sequence alignments, mRNA translation, and amiRNA design. In addition, the minimum hybridization free energy (MFE) between the designed amiRNAs and their corresponding target mRNA sequences was evaluated based on the publicly available RNAhybrid tool (Rehmsmeier et al., RNA (New York, NY), 2004, 10(10): 1507-1517).

[0258] Two groups of multi-lipase targets were defined: (a) LPL and PLBL2 (LP) and (b) LPL and its eight homologs (LH) (Lipase 1a-Lipase 1h). Based on transcript alignment, a seed region shared between lipases was identified and amiRNAs were designed using this seed region, applying features commonly used to predict miRNA target binding. Therefore, all amiRNAs were designed to be 100% complementary to the LPL mRNA sequence. No further preference for amiRNA design against specific lipases was applied. In order to identify suitable amiRNAs capable of knocking down (KD) multiple lipases.

[0259] Table 5: Binding parameters of amiRNAs targeting LPL and PLBL2 and LPL and its homologs.

[0260]

[0261]

[0262] Artificial miRNA design targeting multiple hydrolases / lipases using a single miRNA

[0263] Although the PS-degrading properties of PLBL2 have not been confirmed, this enzyme was included in the multi-lipase KD approach because it is a frequently reported contaminant in antibody products and has the potential to cause HCP-mediated immunogenic responses in patients. Therefore, the amino acid sequences of LPL and PLBL2 were aligned ( Figure 1 B).

[0264] Alignment of LPL and its eight homologs revealed highly conserved amino acid sequence segments among the proteins ( Figure 1 A) These conserved regions are separated by amino acid sequences that have little or no homology between proteins. Figure 1 A highly conserved region was found in the region of amino acids 230 to 270 with consensus numbering shown in A. Regarding the alignment of LPL and PLBL2 at the amino acid sequence level, smaller sequence homologies were identified compared to the alignment of LPL and its relatives ( Figure 1 B) In Figure 1 In , only the amino acid regions are highlighted, showing the high homology between the enzymes.

[0265] For the de novo design of amiRNA, further comparison at the transcriptional level was performed ( Figure 2 ). Based on the design criteria presented in Table 4, manual de novo amiRNA design was applied in highly conserved transcript regions.

[0266] The transcript alignments of LPL against its homologs and LPL against PLBL2 are depicted in Figure 2 A and B. Therefore, depending on Figure 2 Homology (lines of identity) in the , conserved nucleotides are depicted as blocks of varying heights. Transcript alignment of LPL and its homologs revealed multiple sites of increased homology separated by stretches that are conserved between the target proteins. In contrast to transcript alignments of LPL and its relatives, homology in the LPL to PLBL2 alignment was found to be more randomly distributed throughout the transcript sequence. Figure 2In the figure, only the transcript regions are highlighted, showing the high homology between the enzymes. The highly conserved regions at the transcriptional level identified by the alignment method were used as target seed regions for designing amiRNAs. Three amiRNAs were designed for LPL and its eight homologs (LH): LH 9mer, non-classical LH, and LH 8mer. In addition, two amiRNAs were designed to bind LPL and PLBL2 (LP): LP 6mer and LP 8mer.

[0267] Figure 3 (LPL and its homologues) and Figure 4 A more detailed route for amiRNA design targeting multiple lipases simultaneously is given in (LPL and PLBL2). Thus, the theoretical amiRNA binding sites in different lipases are highlighted. For multi-lipase amiRNA design, LPL was given priority as the primary target due to its high potential to degrade PS. Therefore, the amiRNA was designed to be close to 100% complementary to the LPL target enzyme (LH 9mer has only 95% complementarity). However, vacancies were allowed for amiRNA binding to LPL. Furthermore, the amiRNA design was applied in a way that it attempted to achieve as much complementarity as possible with respect to the remaining lipase targets. Therefore, in Figure 3 and 4 In the figure, conserved nucleotides between targets are depicted as black boxes, while non-identical nucleotides are depicted as dots, and gaps between mRNA targets and de novo designed amiRNAs are marked with dashes ( Figure 4 B). The detailed design and binding parameters of all amiRNAs are summarized in Table 5. In brief, 22-23 bp long amiRNAs with different seed lengths of 5-9 nucleotides and GC contents of 45-61% were designed. In addition, the 3' end compensation of LP amiRNA was kept above 47%. The MFE values ​​describing the binding energy between amiRNA and mRNA were generally lower than -26 kcal / mol. The total number of matching nucleotides for LP amiRNA was 16-17 bp, while that for LH amiRNA was 9-10 bp. Except for LP 8mer, none of the theoretical amiRNA binding sites showed a bulge (at Figure 4 In the LP 8mer binding, both targeted mRNA sequences exhibited an A at position 1, whereas in the LP 6mer binding site, no A was found at position 1 of the mRNA. For the LH amiRNAs, 22-66% of the targeted mRNAs exhibited an A at mRNA position 1. Of the eight homologs in total, the LH amiRNAs exhibited perfect seed matches to three to five homologs.

[0268] Regarding amiRNA design against LPL and its homologs, all amiRNA binding sites were generated with a preference for optimal seed binding across the target ( Figure 3 ). Therefore, more amiRNA complementary nucleotides are obtained at the 5' end of the amiRNA sequence (black box). The main differences between the designed LH amiRNAs are the target region and the corresponding seed length. LH 9mer is designed to bind to nine nucleotide seed lengths ( Figure 3 A). Thus, 100% seed complementarity was achieved for the transcripts of lipase 1a, lipase 1b and lipase 1d. For all remaining lipase targets, mismatched amiRNA:mRNA binding in the 9mer seed region was obtained. For non-classical LH, an 11 nucleotide long seed region was defined ( Figure 3 B). However, for all lipases except lipase 1g, LH non-classical amiRNAs did not show seed binding of more than six consecutive complementary nucleotides. Similar to LH 9mer, LH 8mer amiRNAs were created with at least 8 complementary nucleotides in the seed region ( Figure 3 C). Thus, 100% seed complementarity was established for the transcripts of lipase 1a, lipase 1d and lipase 1f. Regarding the remaining targets, target binding in the seed region was achieved with 1-2 mismatches between the amiRNA and the corresponding mRNA. In addition, the binding capacity of the amiRNA to its target was increased by introducing a 3' end compensation in the amiRNA sequence of complementary nucleotides shared between the mRNA and the de novo miRNA molecule. Compensation at the 3' end of the amiRNA was performed for all LH RNA molecules.

[0269] The binding sites of amiRNAs targeting LPL and PLBL2 are shown in Figure 4 Similar to LH amiRNA, LP amiRNA design prioritizes LPL. The two LP amiRNAs were designed to bind in different regions of the lipase. In addition, during the de novo design of LPamiRNA, seed length variation and 3' end target binding compensation were applied. In addition to the six complementary nucleotides in the seed region, the LP 6mer was also designed to bind to PLBL2 via six consecutive nucleotides at the 3' end of the amiRNA sequence ( Figure 4 A). In contrast, LP 8mer was created with eight complementary nucleotides in the seed region. However, the compensation at the 3' end of the amiRNA was not as strong as that of LP 6mer ( Figure 4 B). In addition, the LP 8mer amiRNA was designed to have two gaps (middle and 3' end of the amiRNA sequence) for PLBL2 target binding (marked with a hyphen (-) in the alignment), resulting in one nucleotide bulge at those sites.

[0270] Overall, three miRNAs were designed against LPL and its homologues, and two de novo RNA molecules were generated that simultaneously targeted LPL and PLBL2. The designed amiRNAs were further processed to evaluate their ability to target multiple lipases in parallel.

[0271] Example 2: Effects of artificial miRNA after transient transfection

[0272] Transient transfection of artificial miRNA targeting lipase in CHO host cells

[0273] As mentioned above, lipases are identified as key CHO host cell impurities causing PS degradation in several DPs. In addition, some difficult-to-remove lipases are found to be immunogenic in patients. In order to reduce difficult-to-remove lipases in the early steps of biopharmaceutical manufacturing, amiRNAs are designed to knock down multiple lipases at the same time. From a long-term perspective, the stable co-expression of amiRNA in CHO production cell lines is the goal of multi-lipase gene regulation. However, the process of stable cell line generation is extremely time-consuming and lengthy. Therefore, the function of amiRNA designed from scratch for multiple lipases is elucidated via transient transfection of amiRNA mimics. Therefore, amiRNAs for LPL and its homologs as well as LPL and PLBL2 are transiently transfected into CHO host cells to evaluate their post-transcriptional gene regulation properties. Subsequently, CHO cell pellets and cell culture supernatants are harvested to evaluate the effects of amiRNA transfection on cell growth, viability, lipase KD and PS degradation.

[0274] Analysis of cell growth and viability

[0275] De novo designed amiRNAs for LH and LP were ordered from ThermoFisher as mirVana mimics and transiently transfected into the CHO-K1 GS host cell line. After transfection, cells were cultured in batch cultures in 96 half-DW microplates. Three days after transfection, cell growth and viability of amiRNA-transfected CHO host cells were determined by flow cytometry ( Figure 5 In addition, HCCF and cell pellet samples were harvested for lipase concentration analysis and determination of PS degradation.

[0276] Figure 5 The effects of amiRNA transfection on cell growth and viability in CHO host cells are presented. On average, except for the fact that the average VCC was 5.0 × 10 6 cells / ml and 0.1×10 6 Except for LH 8mer and Tox5, most transfections achieved 6.5-8.0×10 6The results show that the VCC of 10 cells / ml was 0.177 cells / ml. Tox5 represents a functional positive control miRNA that induces cell death after transfection. Regarding viability, all amiRNAs and control transfections measured values ​​of >97% except Tox5. As expected, cells transfected with Tox5 resulted in an average viability of 16%. Overall, especially for LH 8mer, a slight effect of multi-lipase amiRNA on cell growth was observed. However, the reduction in VCC values ​​of samples transfected with multi-lipase amiRNA was not significant.

[0277] In subsequent experiments (data not shown) using more replicates (n=4), significant differences in the activity of LH 8mer and non-classical LH were observed. However, these changes were negligible as the absolute differences were in the range of 1%.

[0278] Determination of lipase concentration

[0279] To evaluate the LPL KD properties of the amiRNAs, LPL (lipase 1) concentrations in HCCF samples transfected with CHO host cells were determined by sandwich ELISA ( Figure 6 ). All amiRNAs were designed to have 100% sequence complementarity to LPL.

[0280] Except for LP 8mer, which resulted in an average normalized LPL concentration of 4.4fg / cell, all multi-lipase targeted amiRNA transfections resulted in an average normalized LPL concentration of <3.0fg / cell. These concentrations translated into a >50% downregulation compared to the NT siRNA control. The LPL siRNA positive control resulted in a significant LPL KD with an average normalized LPL concentration of 0.9fg / cell relative to the 5.6fg / cell measured by the NT siRNA control. For the mock control, an average normalized LPL concentration of 5.5fg / cell was measured, which was within the range of the NT siRNA sample. For most of the de novo designed multi-lipase amiRNAs, a clear trend of post-transcriptional downregulation of LPL was detected.

[0281] In subsequent experiments with more replicates (n=4) (data not shown), all miRNAs triggered a significant reduction in LPL abundance >50%, with mean normalized LPL concentrations <3.0 fg / cell, with the exception of LP 8mer, which reduced LPL expression by only 33% compared to NT controls (<4.0 fg / cell).

[0282] To investigate whether artificial amiRNA mimics could downregulate not only LPL but also PLBL2 (lipase 2) expression, HCCF samples of amiRNA-transfected CHO host cells were subjected to PLBL2 biolayer interferometry (BLI) assay to quantify PLBL2 concentration ( Figure 7 ). The LP amiRNA was specifically designed to target PLBL2 mRNA in addition to LPL mRNA, while the LH amiRNA was constructed to target LPL and its eight homologues. However, all amiRNA transfected samples were analyzed in the PLBL2 BLI assay.

[0283] LP 8mer, LH 8mer and LH non-classical amiRNAs did not show a significant reduction in PLBL2 expression, resulting in normalized PLBL2 concentrations ranging from 23.5 to 25.9 fg / cell. These measured concentrations are consistent with the values ​​generated by the NT siRNA reference control, which had a normalized PLBL2 concentration of an average of 26.0 fg / cell. Relative to the NT siRNA control, significantly reduced PLBL2 concentrations were measured in HCCF samples of CHO host cells transfected with LH 9mer and LP 6mer - normalized PLBL2 concentrations of 14.6 and 18.2 fg / cell, respectively. Therefore, transfection of LH 9mer and LP 6meramiRNAs resulted in a 44% and 30% reduction in normalized PLBL2 concentrations relative to the NT siRNA control.

[0284] All amiRNAs were designed with the goal of targeting multiple lipases. In particular, LH amiRNAs should target LPL and its various homologs. In order to study the amiRNA KD properties with respect to other lipases, Pla1a (lipase 1a) was selected as a representative homolog of LPL. Therefore, sandwich ELISA was used to quantify Pla1a. Due to the fact that Pla1a was found to be intracellularly localized in cells, cell lysis was performed before Pla1a quantification. Therefore, CHO host cells were harvested 3 days after transfection with amiRNAs, and Pla1a in the cell lysate was quantified. In addition, the total protein content of the cell lysate was measured, so that the Pla1a concentration could be normalized based on the total protein content ( Figure 8 ). Although LP amiRNA was not designed to target Pla1a, samples transfected with LP amiRNA were included in the analysis to evaluate the potential off-target KD properties of LP amiRNA for Pla1a.

[0285] like Figure 8As shown, normalized Pla1a concentrations were reduced in cell lysate samples of LH amiRNA transfected CHO host cell lines (7.3-8.0 ng / mg) compared to normalized Pla1a concentrations measured in NT and mock controls, respectively (9.4 and 9.7 ng / mg). Thus, significantly reduced Pla1a concentrations (7.4 ng / mg on average) were measured in cell lysate samples transfected with non-classical LH relative to NT controls. Notably, LP amiRNA mimics resulted in reduced Pla1a concentrations - 8.1 ng / mg for LP 6mer transfected samples and 5.9 ng / mg for LP 8mer amiRNA transfection. On average, a 14-37% reduction in Pla1a protein concentration was detected in cell lysate samples transfected with lipase-targeted amiRNAs.

[0286] Strong downregulation of Pla1a was observed after LP 8mer transfection, although this amiRNA was not designed to target LPL and its homologues ( Figure 8 Similar observations were made for LH 9mers specifically designed to bind to transcript sequences of LPL and its homologs. Strikingly, in addition to post-transcriptional regulation of LPL and Pla1a, transient application of LH 9mers in CHO host cells also resulted in a strong downregulation of PLBL2 ( Figure 7 To identify the underlying cause of these unexpected observations, off-target analyses were performed to analyze potential binding sites of the LP 8mer and LH 9mer amiRNAs in the Pla1a transcript sequence and PLBL2 transcript sequences, respectively ( Fig. 9 ). Therefore, less stringent settings were applied for amiRNA:mRNA alignments.

[0287] For LP 8mer amiRNA, the theoretical binding site in the transcript sequence of Pla1a was identified to have eight mismatches between mRNA and amiRNA. At the 5' end of amiRNA, partial sequence complementarity of four nucleotides was observed. In addition, in the middle of the mRNA:amiRNA comparison, three matching nucleotides were identified, which were separated by mismatches. In addition, a section of six complementary nucleotides was found at the 3' end of amiRNA. For LH 9mer amiRNA, the theoretical binding site in the transcript sequence of PLBL2 was identified to have only seven mismatches between mRNA and amiRNA. At the 5' end and towards the middle of amiRNA, three sections of four complementary nucleotides were observed. In addition, three single complementary nucleotides were found at the 3' end of amiRNA. In summary, the potential off-target binding sites of LP 8mer and LH 6mer amiRNA in the lipase transcript sequence (these sites are not intended to be targeted by one of the corresponding amiRNAs) were identified as the root cause of nonspecific lipase downregulation. The results were further confirmed in cells stably expressing mAbl (IgGl) or mAb2 (IgG4), which further showed no effect on volumetric and specific productivity, with the exception of LH 9mer which showed an increase in specific productivity. A summary of the results is provided in Table 6.

[0288] Table 6: Summary of the effects of de novo designed amiRNAs

[0289]

[0290] Determination of polysorbate degradation

[0291] In order to evaluate the potential of de novo designed amiRNA to downregulate multiple lipases and thus the effect on PS degradation, HCCF samples of amiRNA-transfected CHO host cells were subjected to FMA analysis, which enabled the quantification of PS degradation over time. Therefore, HCCF samples were diluted 1:50 with medium spiked with PS20 (c = 0.4 mg / ml) and PS80 (c = 0.2 mg / ml) and incubated in the dark at room temperature (about 22 ° C) for up to 14 days. Subsequently, after adding a surfactant, the PS-spiked samples were stored for 0, 1, 3, 7 and 14 days. After the incubation period, the PS concentration was measured via micelle-embedded fluorescent molecules that can be bound to PS micelles. Compared with the NT siRNA control, the degradation of PS in amiRNA-transfected CHO host cell samples was evaluated. In addition, the mimic transfection control of the buffer and the LPL-targeted siRNA control were included in the FMA study. The resulting PS concentrations at different time points are depicted in Figure 10.

[0292] Negative controls transfected with NT siRNA or nuclease-free water (simulations) showed strong PS20 and PS80 degradation within 14 days of incubation. After 14 days of incubation after PS addition, approximately 0.13 mg / ml PS20 and 0.005 mg / ml PS80 were measured in two samples, respectively. From the 3rd day, compared with the NT siRNA control, PS20 and PS80 degradation were detected to be significantly reduced in the CHO samples transfected with amiRNA. In the HCCF samples transfected with LP 6mer, LH 8mer, LH 9mer and LH non-classical amiRNA, it was observed that the PS degradation of PS20 and PS80 was quite. Therefore, in these samples incubated for 14 days, the PS20 concentration was measured to be between 0.31 and 0.34 mg / ml, and the PS80 concentration was in the range of 0.062 to 0.081 mg / ml. Thus, relative to the initial PS concentration, 21-30% of PS20 in these samples degraded over time, while 68% of PS20 in the NT sample degraded. During the 14-day incubation period, 63-71% of the initial PS80 concentration decreased over time in the HCCF samples spiked with PS80 transfected with LP 6mer, LH 8mer, LH 9mer, and LH non-classical amiRNA, while PS80 decreased by 98% in the NT siRNA control sample. In addition, lower PS concentrations were determined in the spiked HCCF samples transfected with LP 8mer compared to the samples of the remaining amiRNA, indicating stronger PS degradation. Thus, after 14 days of incubation, 0.21 mg / ml PS20 and 0.009 mg / ml PS80 were determined in the spiked LP 8mer HCCF samples, respectively. However, at the end of the FMA incubation, the measured PS concentrations were significantly higher than those detected in the NT siRNA control. Compared to all the remaining samples, HCCF samples from CHO host cells transfected with LPL siRNA showed the least PS20 and PS80 degradation. For PS20, a concentration of 0.39 mg / ml was measured at the 14th day of the FMA incubation time point, thus, only 7% of the initial PS20 concentration was degraded. A PS80 concentration of 0.168 mg / ml was determined in these samples, indicating a 21% decrease in the initial PS concentration.

[0293] In addition, the average degradation rates of PS20 and PS80 in HCCF samples were calculated ( Fig.11 ). Therefore, the difference between the initial (day 0) and the remaining PS concentration (day 14) was divided by 14 days to obtain the daily PS degradation rate (in mg / ml*day).

[0294] In general, except for the mimic control, all HCCF samples showed a significant reduction in PS degradation rate relative to the NT siRNA control. Thus, in the NT siRNA sample, 0.020 mg / ml PS20 and 0.015 mg / ml PS80 were degraded every day. In the mimic HCCF sample, no changes in the two PS degradation rates were measured relative to the Scramble control. LP 6mer and all LH transfections resulted in similar degradation rates in the spiked HCCF samples-0.007-0.009 mg / ml*day was observed for PS20 and 0.009-0.011 mg / ml*day was observed for PS80. In addition, higher PS degradation rates were measured in the spiked HCCF samples of LP 8mer transfection compared to the samples of the remaining amiRNA. Therefore, in the spiked LP 8merHCCF samples, 0.016 mg / ml PS20 and 0.015 mg / ml PS80 were degraded every day. The lowest PS degradation rate was observed in the LPL siRNA control HCCF samples. For PS20, a degradation rate of 0.002 mg / ml*day was calculated, and for PS80, a degradation rate of 0.003 mg / ml*day was calculated. In general, in HCCF samples where LPL KD>50% was observed, PS80 degradation rates were higher than PS20 degradation rates ( Fig.11 ).

[0295] Example 3: Function of amiRNA in IgG-producing CHO cell lines

[0296] Considering the stable overexpression of amiRNAs in CHO production cell lines, it is critical to ensure that the designed amiRNAs do not impair cell growth, viability, or productivity of protein therapeutics. Therefore, representative IgG1 and IgG4 CHO-K1 GS were transiently transfected with anti-lipase amiRNAs. - / - Production cell lines and analysis of the effects of de novo amiRNA on cell growth, viability and productivity. After transfection, IgG1 and IgG4 production cells were cultured in batches for 5 days. Viability and viable cell concentration (VCC) were measured from day 3 to day 5, and the integral viable cell concentration (IVC) was calculated on day 5. Growth performance data are summarized in Figure 12.

[0297] The introduction of LH 9mer amiRNA resulted in a significant decrease in the viability of IgG1 and IgG4 producing cells on day 5 to 80.3% and 87.3%, respectively. All other amiRNA-transfected cells (including NT siRNA-transfected cells) showed a sustained high viability of >96% in both IgG-producing CHO cell lines. For IVC, LH 9mer amiRNA transfection resulted in a significant decrease in IVC of both IgG-producing CHO cell lines, while cell growth of cells transfected with LH 8mer amiRNA increased. In both IgG-expressing CHO cell lines, cell growth was not significantly affected by LP 6mer and LP 8mer amiRNAs compared to the NT control. Non-classical LH transfection in IgG1 producing cells did not affect cell growth, while the IVC of IgG4-producing CHO cells transfected with atypical LH was significantly increased compared to the control.

[0298] Finally, the effect of amiRNA transfection on volume and cell specific IgG productivity was evaluated (Figure 13). In IgG1 expressing cell lines, the previously observed growth inhibitory effect of LH 9mer was consistent with a significantly increased specific productivity. However, the volume titer was significantly reduced. Atypical LH transfection resulted in a significant improvement in the final product concentration in IgG1 producing CHO cell lines. All other transfected amiRNAs did not significantly change the volume or specific IgG1 productivity. In IgG4 producing CHO cell lines, it was observed that all transfected amiRNAs had no effect on protein production.

[0299] discuss

[0300] To date, the concept of multi-target knockdown (KD) using rationally designed amiRNAs has remained largely unexplored. The few studies in this area have focused on seed binding in the 3'-UTR while ignoring the remaining conserved regions in the 5'-UTR and coding sequences (Arroyo et al., 2014; Guire et al., 2010), or have altered single nucleotides in natural miRNAs to reduce off-target effects (Klingler et al., 2023).

[0301] Applying this technology for the first time in CHO cells, we were able to identify multi-lipase targeting amiRNAs that simultaneously reduced the lipase expression levels of LPL and PLBL2 or PLA1A. Since the conceptualization of this work, although PLBL2 is highly abundant in medicines, it has been shown that PLBL2 is not the cause of PS degradation (S. Zhang et al., 2020, Journal of Pharmaceutical Sciences, 109 (11), 3300-3307). However, as a lipase that is difficult to remove that has been shown to induce immune responses in mAb medicines, it is an effective model protein that should be removed (SK Fischer et al., 2017, The AAPS Journal, 19 (1), 254-263). In addition to strong evidence of multiple lipase KD at the protein level (which makes evidence at the transcriptional level redundant), FMA verifies KD at the functional level. The supernatant of CHO cells transfected with amiRNA spiked with PS is significantly slower than the control sample in degradation of PS substrates. Furthermore, stronger PS degradation was observed for PS80, indicating that LPL / PLA1A preferentially degrades PS20.

[0302] By carefully evaluating different amiRNAs, it may even be possible to identify additional pro-production effects of amiRNA transfection (see non-classical LH in IgG1 producing cells), which has been used in the past to engineer CHO cell lines with endogenous miRNAs (Barron et al., 2011, Journal of Biotechnology 151(2), 204-211; S. Fischer et al., 2017, Biotechnology and Bioengineering, 114(7), 1495-1510; S. Fischer et al., 2015, Journal of Biotechnology, 212, 32-43; Loh et al., 2014, Biotechnology Journal, 9(9), 1140-1151; Loh et al., 2017, Biotechnology Journal, 12(4), https: / / doi.org / 10.1002 / biot.201600488).

[0303] The present examples provide the first proof of concept that multi-target regulatory amiRNAs can be rationally designed and generated and represent a powerful and flexible tool for CHO cell line engineering. They were successfully applied to mitigate PS degradation in CHO cell culture supernatants by reducing the expression of key HCPs. Considering that all experiments were performed transiently, stable expression of amiRNAs is expected to have an even more pronounced effect on PS degradation, as transient transfection may not be able to effectively reduce the expression of HCPs with longer half-lives and amiRNAs will be degraded relatively quickly over time. In the future, careful amiRNA design will be able to target an even larger class of mammalian hydrolases that act on ester bonds and exhibit "micro-conserved" regions in their mRNA sequences. Other conceivable applications in order to improve the product quality of the therapeutic protein of interest are to reduce immunogenic HCPs, such as PLBL2.

Claims

1. An artificial microRNA (miRNA) comprising an artificial miRNA guide strand sequence, wherein the artificial miRNA guide strand sequence targets two or more mammalian hydrolases acting on ester bonds.

2. The artificial miRNA according to claim 1, wherein (a) the two or more hydrolases are endogenous to mammalian cells, preferably human or rodent cells, more preferably CHO cells; and / or (b) wherein the artificial miRNA guide strand sequence has complete or incomplete complementarity with the target sequences of at least two mRNAs encoding the two or more mammalian hydrolases; and / or (c) wherein the target sequence is located in the 3' untranslated region or the coding region of at least two mRNAs encoding the two or more mammalian hydrolases.

3. The artificial miRNA according to claim 1 or 2, wherein the artificial miRNA targeting two or more mammalian hydrolases reduces the activity of the two or more hydrolases in mammalian cells, preferably rodent or human cells, more preferably CHO cells.

4. The artificial miRNA according to any one of claims 1 to 3, wherein (a) the artificial miRNA guide strand sequence has 19 to 25 nucleotides, preferably 21 to 23 nucleotides; and / or (b) the artificial miRNA guide strand sequence independently hybridizes with the target sequences of at least two mRNAs encoding the two or more mammalian hydrolases without forming a bulge loop and / or forming at most two bulge loops; and / or (c) the artificial miRNA guide strand sequence has at least 55% sequence complementarity with one or more target sequences of at least two mRNAs encoding the two or more mammalian hydrolases; and / or wherein the miRNA guide strand sequence has at least 13 Watson-Crick base pairings with one or more target sequences of at least two mRNAs encoding the two or more mammalian hydrolases.

5. The artificial miRNA according to any one of claims 1 to 4, wherein (a) one of the two or more mammalian hydrolases is lipoprotein lipase (LPL); and / or (b) one of the two or more mammalian hydrolases is LPL, and the one or more target sequences in the mRNA encoding LPL are within nucleotides 560-590, 700-745, 930-960 or 1330-1360 of SEQ ID NO: 6 or 7, preferably within nucleotides 560-590, 705-745, or 930-960 of SEQ ID NO: 6 or 7; and / or (c) the two or more mammalian hydrolases are selected from lipoprotein lipase (LPL), phospholipase B-like protein 2 (PLBL2), phospholipase A1 (Pla1a), inactive pancreatic lipase-related protein 1 (PNLIPRP1), inactive pancreatic lipase-related protein 2 (PNLIPRP2), pancreatic triacylglycerol lipase precursor (PNLIP), lipase member I precursor (LipI), lipase member H precursor (LipH), lipase member G precursor (LipG), lipase member C precursor (LipC) and isoforms thereof, and / or (d) wherein the artificial miRNA guide strand sequence has a nucleotide sequence of SEQ ID NO: 1 (LP 8mer), SEQ ID NO: 2 (LP6mer), SEQ ID NO: 3 (LH 8mer), SEQ ID NO: 4 (LH 9mer) or SEQ ID NO: 5 (non-classical LH).

6. The artificial miRNA according to any one of the preceding claims, wherein the artificial miRNA guide strand sequence has incomplete complementarity with one or more target sequences of at least one of the at least two mRNAs encoding the two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) a sequence of 8 nucleotides (nt 2-9) which is complementary to the one or more target sequences and has 0, 1, 2 or 3 mismatches, preferably 0, 1 or 2 mismatches, with the one or more target sequences, and (ii) a GC content of 55-70%, wherein all guanines (G) or cytosines (C), or all guanines (G) or cytosines (C) except one guanine (G) or cytosine (C), or all guanines (G) or cytosines (C) except two guanines (G) or cytosines (C) are complementary to the one or more target sequences; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides which is complementary to the one or more target sequences and has 1, 2 or 3 mismatches with the one or more target sequences, preferably wherein the mismatches are not consecutive mismatches, and (ii) The GC content of the last 6 nucleotides is ≤50%, wherein 1 or 2 complementary nucleotides are G or C.

7. The artificial miRNA according to claim 6, wherein the artificial miRNA guide strand sequence has the nucleotide sequence of SEQ ID NO: 2 (LP 6mer), SEQ ID NO: 4 (LH 9mer) or SEQ ID NO: 5 (non-classical LH).

8. The artificial miRNA according to any one of claims 1 to 5, wherein the artificial miRNA guide strand sequence has incomplete complementarity with one or more target sequences of at least one of the at least two mRNAs encoding the two or more mammalian hydrolases, and wherein the miRNA guide strand sequence comprises a 5' seed region starting at nucleotide 2 (nt 2) and a 3' complementary region that undergoes base pairing with the one or more target sequences, wherein the 5' seed region comprises (i) a sequence of 8 nucleotides (nt 2-9), wherein at least the first 4 consecutive nucleotides (at least nt 2-5) are complementary to the one or more target sequences, wherein the sequence of 8 nucleotides comprises at least one mismatch with the one or more target sequences, and (ii) a GC content of 55-75%, wherein at least 4 consecutive nucleotides complementary to the one or more target sequences are G or C; and wherein the 3' complementary region comprises (i) a sequence of the last 6 nucleotides which is complementary to the one or more target sequences and has 0 or 1 mismatches with the one or more target sequences, preferably wherein the last nucleotide is complementary to the one or more target sequences and / or the second to last nucleotide is a mismatch, and (ii) The GC content of the last 6 nucleotides is 30-50%, and preferably at least 2 complementary nucleotides are G or C. 9 . The artificial miRNA according to claim 8 , wherein the artificial miRNA guide strand sequence has the nucleotide sequence of SEQ ID NO: 1 (LP 8mer) or SEQ ID NO: 2 (LP 6mer).

10. The artificial miRNA according to any one of claims 1 to 9, wherein the artificial miRNA is a pri-miRNA, a pre-miRNA, a double-stranded mature miRNA (comprising a follower strand and a guide strand), a single-stranded miRNA guide strand, or a single-stranded or double-stranded miRNA mimic.

11. A mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA according to any one of claims 1 to 10.

12. The mammalian expression vector according to claim 11, wherein (a) the mammalian expression vector further comprises at least one target gene encoding a target recombinant protein; and / or (b) the mammalian expression vector comprises a polynucleotide sequence encoding two or more artificial miRNAs according to any one of claims 1 to 10, wherein the artificial miRNAs are the same or different.

13. A mammalian cell comprising the artificial miRNA according to any one of claims 1 to 11 or the mammalian expression vector according to claim 11 or 12.

14. The mammalian cell according to claim 13, wherein (a) the mammalian cell is a human or rodent cell, preferably a CHO cell; and / or (b) The mammalian cells also express the target recombinant protein.

15. Use of the artificial miRNA according to any one of claims 1 to 10 in a mammalian cell for preventing the expression of at least two endogenous hydrolases in the mammalian cell and / or preventing the hydrolase activity and / or polysorbate degradation activity associated with a recombinant protein of interest produced in the mammalian cell.

16. A method for producing a target protein, comprising the following steps: (a) introducing an expression vector containing a target gene encoding a target recombinant protein into mammalian cells; (b) introducing an artificial miRNA comprising an artificial miRNA guide strand sequence targeting two or more mammalian hydrolases acting on ester bonds into the mammalian cells, wherein the artificial miRNA is introduced into the mammalian cells before, simultaneously with or after step (a), wherein the artificial miRNA is introduced as RNA or as a mammalian expression vector comprising a polynucleotide sequence encoding the artificial miRNA; (c) culturing the mammalian cells under conditions that allow expression of the recombinant protein of interest; and (d) harvesting the target recombinant protein from the harvested cell culture fluid; as well as Optionally, the following steps are also included: (e) purifying the target recombinant protein; as well as (f) formulating the purified target recombinant protein into a composition.

17. The method according to claim 16, wherein (a) the harvested cell culture fluid has reduced hydrolase activity and / or reduced polysorbate degrading activity; and / or (b) the composition comprising the purified recombinant protein of interest of step (e) or the composition of step (f) has reduced hydrolase activity and / or reduced polysorbate degrading activity; and / or (c) the artificial miRNA prevents the expression of at least two endogenous hydrolases in the mammalian cell, and / or prevents the hydrolase activity and / or polysorbate degradation activity associated with the recombinant protein of interest produced in the mammalian cell.

18. The method according to claim 16 or 17, wherein (a) the artificial miRNA is the artificial miRNA according to any one of claims 1 to 10; and / or (b) the expression vector comprising the polynucleotide sequence encoding the artificial miRNA is a mammalian expression vector according to claim 11 or 12; and / or (c) the two or more mammalian hydrolases are endogenous to the mammalian cell, preferably the mammalian cell is a human or rodent cell, more preferably a CHO cell; and / or (d) wherein the artificial miRNA is introduced as RNA by transient transfection, or as an expression vector comprising a polynucleotide sequence encoding the artificial miRNA by transfection or transduction.

Images