Sequencing the CHO DXB11 genome reveals regional variations in genomic stability and haploidy.

BACKGROUND: The DHFR negative CHO DXB11 cell line (also known as DUX-B11 and DUKX) was historically the first CHO cell line to be used for large scale production of heterologous proteins and is still used for production of a number of complex proteins. RESULTS: Here we present the genomic sequence of the CHO DXB11 genome sequenced to a depth of 33x. Overall a significant genomic drift was seen favoring GC → AT point mutations in line with the chemical mutagenesis strategy used for generation of the cell line. The sequencing depth for each gene in the genome revealed distinct peaks at sequencing depths of 0x, 16x, 33x and 49x coverage corresponding to a copy number in the genome of 0, 1, 2 and 3 copies. This indicate that 17% of the genes are haploid revealing a large number of genes which can be knocked out with relative ease. This tendency of haploidy was furthermore shown to be present in eight additional analyzed CHO genomes (15-20% haploidy) but not in the genome of the Chinese hamster. The dhfr gene is confirmed to be haploid in CHO DXB11; transcriptionally active and the remaining allele contains a G410C point mutation causing a Thr137Arg missense mutation. We find ~2.5 million single nucleotide polymorphisms (SNP's), 44 gene deletions in the CHO DXB11 genome and 9357 SNP's, which interfere with the coding regions of 3458 genes. Copy number variations for nine CHO genomes were mapped to the chromosomes of the Chinese hamster showing unique signatures for each chromosome. The data indicate that chromosome one and four appear to be more stable over the course of the CHO evolution compared to the other chromosomes thus might presenting the most attractive landing platforms for knock-ins of heterologous genes. CONCLUSIONS: Our studies reveal an unexpected degree of haploidy in CHO DXB11 and CHO cells in general and highlight the chromosomal changes that have occurred among the CHO cell lines sequenced to date.

Network reconstruction of the mouse secretory pathway applied on CHO cell transcriptome data.

BACKGROUND: Protein secretion is one of the most important processes in eukaryotes. It is based on a highly complex machinery involving numerous proteins in several cellular compartments. The elucidation of the cell biology of the secretory machinery is of great importance, as it drives protein expression for biopharmaceutical industry, a 140 billion USD global market. However, the complexity of secretory process is difficult to describe using a simple reductionist approach, and therefore a promising avenue is to employ the tools of systems biology. RESULTS: On the basis of manual curation of the literature on the yeast, human, and mouse secretory pathway, we have compiled a comprehensive catalogue of characterized proteins with functional annotation and their interconnectivity. Thus we have established the most elaborate reconstruction (RECON) of the functional secretion pathway network to date, counting 801 different components in mouse. By employing our mouse RECON to the CHO-K1 genome in a comparative genomic approach, we could reconstruct the protein secretory pathway of CHO cells counting 764 CHO components. This RECON furthermore facilitated the development of three alternative methods to study protein secretion through graphical visualizations of omics data. We have demonstrated the use of these methods to identify potential new and known targets for engineering improved growth and IgG production, as well as the general observation that CHO cells seem to have less strict transcriptional regulation of protein secretion than healthy mouse cells. CONCLUSIONS: The RECON of the secretory pathway represents a strong tool for interpretation of data related to protein secretion as illustrated with transcriptomic data of Chinese Hamster Ovary (CHO) cells, the main platform for mammalian protein production.