Identification of transgene integration loci of different highly expressing recombinant CHO cell lines by FISH

Abstract

Recombinant CHO cell lines have integrated the expression vectors in various parts of the genome leading to different levels of gene amplification, productivity and stability of protein expression. Identification of insertion sites where gene amplification is possible and the transcription rate is high may lead to systems of site-directed integration and will significantly reduce the process for the generation of stably and highly expressing recombinant cell lines. We have investigated a broad range of recombinant cell lines by FISH analysis and Giemsa-Trypsin banding and analysed their integration loci with regard to the extent of methotrexate pressure, transfection methods, promoters and protein productivities. To summarise, we found that the majority of our high producing recombinant CHO cell lines had integrated the expression construct on a larger chromosome of the genome. Furthermore, except from two cell lines, the exogene was integrated at a single site. The dhfr selection marker was co-localised to the target gene.

Fig. 1

Hybridisation patterns of the EpoFc clones: 2C10 (a), 2G6 (b), 2C10/13F5 (c), 2G4 (d), HI5 (e), HI5/7D3 (f) and 10D9 (g). Integration sites are marked with a white arrow

Fig. 2

Giemsa–Trypsin banding patterns of selected EpoFc clones compared to the host cell line (a): EpoFc 2C10 (b) and EpoFc 10D9 (c). The site of integration in EpoFc 2C10 is marked with an arrow, whereas it could not be identified in EpoFc 10D9 since the signal on the short arm of an acrocentric chromosome could not be related to a certain chromosome. FISH-analysis on banded chromosomes might give further information about the integration locus

Fig. 3

Southern blot after digestion of 1.5 μg of genomic DNA for EpoFc clones (4 μg for clone EpoFc HI5/7D3): M: λ DNA/EcoRI + HindIII marker; further lanes show the genomic DNA of the clones. Intensities correlate well with gene copy numbers determined by quantitative PCR, but not with signal intensities detected in FISH analyses

Fig. 4

Hybridisation patterns of the following clones: IgG 2F5 (a), IgG 2F5 sf (b), IgG 2G12 (c), IgG 4E10 (d), IgM Hb617 (e), Fab 3H6 (f), HGP (g). Integration sites are marked with a white arrow

Fig. 5

Giemsa–Trypsin banding patterns of the clones compared to the host cell line (a): IgG 2F5 (b) and HGP (c). The sites of integration are marked with an arrow