
One at a time is not an efficient strategy for hot dog-eating champions—or for knocking out every gene in the mouse genome. Gragerov et al. report a new high-throughput knockout technique based on retroviruses in a recent article in Proceedings of the National Academy of Sciences.
Researchers can induce genome-wide point mutations with chemical mutagenesis. However, point mutations rarely inactivate genes. In insertional mutagenesis, exogenous vectors integrate randomly into the host genome and disrupt gene expression.
Retroviruses insert their genomes randomly into the host genome. The authors designed and patented a retroviral vector based on Moloney murine leukemia virus. The vector contained translation stop codons in all reading frames and a transcription terminator from the human complement gene to block target gene expression. The vector also contained a splice acceptor site, ensuring that it would not be spliced out during RNA processing if it integrated into an intron.
The authors infected mouse embryonic stem (ES) cells with the retroviral vector, generating approximately 10 million different ES cell clones. They pooled 500 clones together in each well of 96-well plates. To detect specific mutants, the authors did PCR with forward primers complementary to the gene of interest and reverse primers complementary to the retroviral vector. They diluted their clone pools until they isolated single clones with the desired mutation. Theoretically, each ES clone could contain many retroviral insertions; however, most clones contained one or two insertions. According to the authors, when generating mutant mice from ES cells with the desired mutation, selective breeding should select for only one mutation.
In the hypoxanthine-guanine phosphoribosyltransferase (HPRT) locus, which is associated with Lesch-Nyhan syndrome, the authors identified 33 different vector integration sites. The authors used an ES cell line derived from males, and the HPRT locus localizes to the X chromosome. Therefore, insertion of the retroviral vector into the HPRT locus should cause HPRT deficiency, which confers resistance to the chemotherapy agent 6-thioguanine. Most of the ES cell clones with vector insertions in HPRT were 6-thioguanine-resistant, consistent with retroviral disruption of gene expression.
Most G protein-coupled receptor genes are difficult targets for insertional mutagenesis because they are small. However, the authors found retroviral insertions in 319 of 355 nonchemosensory G protein-coupled receptor genes. Although researchers believed that retroviruses preferentially affected active genes, the authors detected insertions in genes that were inactive in ES cells.
The authors generated mutant mice from 60 ES cells carrying mutations in G protein-coupled receptors and bred them to homozygosity. Of the 60 resulting mutant mouse lines, 57 lacked the appropriate gene product as desired. Vector insertion upstream of target gene coding regions did not block target gene expression.
Using the authors' technique, many knockout mice could be made in parallel. Because the retroviral vector also contains loxP recombination sites and the tetracycline transactivator, the vector could alternatively be used to inducibly express exogenous gene products from many different loci in the mouse genome.
