Recombinant protein expression by targeting pre-selected chromosomal loci

Abstract

Background: Recombinant protein expression in mammalian cells is mostly achieved by stable integration of transgenes into the chromosomal DNA of established cell lines. The chromosomal surroundings have strong influences on the expression of transgenes. The exploitation of defined loci by targeting expression constructs with different regulatory elements is an approach to design high level expression systems. Further, this allows to evaluate the impact of chromosomal surroundings on distinct vector constructs.

Results: We explored antibody expression upon targeting diverse expression constructs into previously tagged loci in CHO-K1 and HEK293 cells that exhibit high reporter gene expression. These loci were selected by random transfer of reporter cassettes and subsequent screening. Both, retroviral infection and plasmid transfection with eGFP or antibody expression cassettes were employed for tagging. The tagged cell clones were screened for expression and single copy integration. Cell clones producing > 20 pg/cell in 24 hours could be identified. Selected integration sites that had been flanked with heterologous recombinase target sites (FRTs) were targeted by Flp recombinase mediated cassette exchange (RMCE). The results give proof of principle for consistent protein expression upon RMCE. Upon targeting antibody expression cassettes 90-100% of all resulting cell clones showed correct integration. Antibody production was found to be highly consistent within the individual cell clones as expected from their isogenic nature. However, the nature and orientation of expression control elements revealed to be critical. The impact of different promoters was examined with the tag-and-targeting approach. For each of the chosen promoters high expression sites were identified. However, each site supported the chosen promoters to a different extent, indicating that the strength of a particular promoter is dominantly defined by its chromosomal context.

Conclusion: RMCE provides a powerful method to specifically design vectors for optimized gene expression with high accuracy. Upon considering the specific requirements of chromosomal sites this method provides a unique tool to exploit such sites for predictable expression of biotechnologically relevant proteins such as antibodies.

Figure 1

Strategy for the identification of single copy tagged high expressing cell clones for cassette exchange. (A) High expression chromosomal loci were tagged with either retrovirally or plasmid-mediated transduced expression cassettes. These cassettes were flanked by a set of heterospecific recombinase target sites. Tagged cells lines were analyzed for the stable integration of one single copy of the respective tagging vector and screened for high expressing integration loci. The transfer of a targeting vector carrying the same heterospecific recombinase target sites as the tagging vector in presence of the Flp-recombinase leads to a site-specific cassette exchange. The selection for successfully targeted clones was performed by complementing a silent, ATG-defective neomycin resistance gene pre-integrated upon tagging. This renders successfully targeted cells resistant to G418. For this purpose, the incoming targeting vector carried next to its gene of interest (antibody expression unit) a specific sequence (P/I) that facilitates expression of the neomycin resistance gene. (B) The vectors used for tagging are depicted. All tagging cassettes contain a promoter or internal ribosomal entry site for activation of the neomycin resistance gene and are flanked by heterospecific (Fwt-F5) FRT sites. Retroviral tagging was performed as described in [20] with a vector that transduces a bicistronic cassette of eGFP and a hygromycin phosphotransferase/thymidine kinase fusion protein. For plasmidic tagging vectors with different reporter genes (eGFP and/or antibody expression unit) as well as varying promoter elements (SV40/PGK) were employed. All tagging vectors express eGFP, either as a fusion protein with the hygromycin phosphotransferase/thymidine kinase or as a single protein, allowing fluorescence-based screening for the expression of the tagging cassette.

Figure 2

Characterization of the tagged integration sites. (A) GFP expression pattern of single copy HEK293 and CHOK1 clones obtained upon retroviral and plasmidic transfer of the tagging cassette. Representative clones are shown. Green lines: tagged clones. Grey lines represent non-transfected HEK293 and CHOK1 cells, respectively. (B) CHO cells were tagged with an antibody expression construct. The antibody expression level production was evaluated. Only cell clones that produced >2 pcd are shown; the data were obtained from 3-5 replicates. Clones showing single copy integration events (data not shown) are marked by asterisks. Their GFP expression profile is given presented on the right. Green lines: tagged clones; grey lines: non-transfected CHOK1 cells. Nomenclature: HRV-GFP: HEK293 cells retrovirally tagged with eGFP (RV-GFP according to figure 1B); HP-GFP: HEK293 cells plasmid-tagged with eGFP (P-GFP); CRV-GFP: CHO cells retrovirally tagged with eGFP (RV-GFP); CP-AbHTG: CHO cells plasmid- tagged with antibody cassette (PAb-HTG).

Figure 3

Gene expression of cell clones after targeting by cassette exchange. Characterization of HP-GFP and CP-GFP upon targeting. The cell clones show homogeneous loss of GFP expression after targeting with an antibody expression vector as depicted in (B). (B) Molecular characterization of the targeting event by PCR. Amplification of a 550 bp fragment from the daughter cell clones (D1-D6) using primers located at the indicated positions (arrows) is specific for integration of the targeting construct (P: parental tagged cell clone). (C) Homogeneity of the targeted daughter clones. The expression level of daughter clones after cassette exchange is compared with the expression level of the parental clones. Upper left: CP-HTG tagged clone targeted with a SV40 promoter driven GFP cassette. Upper right: HRV-GFP tagged clone targeted with a CMV-RFP cassette. Antibody targeting: CP-HTG tagged clone (lower left) and CP-AbHTG tagged clone (lower right) were targeted with an antibody expression construct as depicted in (B). The expression of the targeted daughter clones is shown.

Figure 4

Antibody expression after targeting in plasmid-mediated and retrovirally tagged HEK293. The retrovirally (RV) or plasmid-mediated (P) tagged GFP expressing HEK293 cells were targeted with the indicated antibody expression cassettes. Antibody expression levels are given as mean expression of 5 targeted daughter clones (pcd). n.t.: not tested.

Figure 5

Antibody expression after targeting of plasmid-mediated tagged CHO-K1. CHO-K1 derived clones plasmid-mediated tagged with either a HTG or AbHTG cassette as indicated were targeted with different antibody expression cassettes. The mean expression of 5 daughter clones after targeting with the depicted constructs as well as the antibody expression of the parental cell in pcd is presented. n.t: not tested.