COMMENTARY

Two ways to trap a in mice

William C. Skarnes* Wellcome Trust Sanger Institute, Hinxton, Cambridge CB10 1SA, United Kingdom

he sequencing of the human tion of random insertional in and mouse genomes has been ES cells (13). Both privately and pub- heralded as one of the most licly funded efforts have generated large important recent achievements libraries of ES cells harboring gene-trap Tin biological science. Attention has now insertions (the International Gene Trap turned to the functional annotation of Consortium maintains a database of the 25,000 or so encoded by the gene-trap cell lines available from the mammalian genome. Although there are public resource at www.genetrap.org). many experimental approaches that ad- Combined, these resources are estimated dress gene function, the most relevant to contain mutations in Ͼ60% of genes approach for extrapolation to human in mice (14). Gene-trap libraries have development, physiology, and disease is become an important source of new mu- to analyze the phenotype of mutations tations in mice for the pharmaceutical for each gene in a whole mammalian industry and academic investigators. model, the mouse. Many phenotypes The National Institutes of Health- relevant to congenital traits and abnor- funded BayGenomics program (http:͞͞ mal pathologies in humans have baygenomics.ucsf.edu), for example, has emerged from gene knockout studies in distributed Ͼ2,000 gene-trap ES cell the mouse, providing a strong justifica- lines to academic laboratories. Clearly, tion for expanding the collection of there is a high demand for mutant ES mouse mutants to include all genes. Re- cell resources within the mouse genetics cent discussions between funding agen- community, and the impact of gene-trap cies and the mouse genetics community resources is beginning to emerge in the (1, 2) have garnered support for an in- published literature (15–18). ternational, concerted effort to generate The targeted trapping approach de- a resource of mutations in every gene in scribed by Friedel et al. (5) represents a the mouse genome. This effort will use Fig. 1. Gene trapping versus targeted trapping. highly efficient method for inactivating a combination of gene targeting and (A) Gene trapping depends on random insertions genes expressed in ES cells. By simply gene trapping in ES cells, two well es- of a promoterless reporter gene, such as ␤geo, flanking a gene-trap cassette with tablished technologies that were devel- equipped with splice acceptor (SA) and polyade- genomic sequences, gene-trap insertions oped in parallel in the late 1980s and nylation (pA) signals. The reporter is activated after can be directed with precision into in- gradually perfected over the years (3, 4). insertions into of expressed genes to gen- trons of genes by homologous recombi- erate a fusion mRNA that can be characterized by 5Ј In this issue of PNAS, Friedel et al. (5) RACE. (B) Targeted trapping relies on homologous nation (Fig. 1B). The only requirement ingeniously combine gene targeting and recombination to introduce a promoterless gene- for the construction of targeted trapping gene trapping to mutate genes expressed trap cassette into predefined loci. The trapping vectors is to avoid including the pro- in ES cells at a high efficiency. It may cassette is flanked by genomic sequences of the moter of the target gene in the arms of surprise many to learn that this hybrid target locus that do not contain the promoter of homology. Like random gene trapping, method of ‘‘targeted trapping’’ is appli- the target gene. activation of the ␤geo-selectable marker cable to a majority of genes in the requires splicing to exons of genes ex- mouse. pressed in ES cells. Thus, correctly tar- Gene targeting has been widely used and increased the precision for building geted events are greatly enriched among over the past 15 years to engineer pre- targeting vectors. Despite these ad- the drug-resistant colonies, and random cise modifications in the mouse genome. vances, gene targeting remains a labor- integrations elsewhere in the genome The collective efforts of many laborato- intensive undertaking that is not easily are suppressed. A further degree of en- ries have thus far produced targeted scalable because of the effort required richment is gained by the use of gene- mutations in Ϸ3,600 genes, or just to screen ES cell colonies for a small trap cassettes that lack a translation Ͻ15% of the genome (a list of targeted fraction of correctly targeted events (9). initiation signal and therefore require gene mutations is maintained by The Gene trapping, by contrast, relies on the production of an in-frame fusion Jackson Laboratory, www.informatics. random integration of a promoterless . The use of promoterless drug jax.org). Gene targeting relies on rare reporter construct (10, 11) and is lim- selection markers for gene targeting is homologous recombination events be- ited to genes expressed in ES cells. The not new (19, 20) but has been used to tween an exogenous DNA construct most widely used vectors contain a target only a handful of genes. Curi- introduced into cells and its cognate splice acceptor and sig- ously, despite the early success of pro- genomic locus, to engineer precise nal flanking the ␤geo reporter gene moterless targeting, this approach was modifications in the genome. Typically, (11), such that the reporter is activated never embraced by researchers. targeting constructs contain several kilo- upon insertion into introns of genes Friedel et al. (5) chose a set of 24 bases of genomic DNA flanking a drug (Fig. 1A). To identify the target gene, genes for targeted trapping, focusing on selection marker driven by a heterologous PCR-based strategies, such as 5Ј RACE, promoter. Improvements in molecular are used to generate a sequence tag for cloning methods, specifically recom- each insertion (12). Gene trapping has See companion article on page 13188. bineering of bacterial artificial chromo- gained prominence in recent years as a *E-mail: [email protected]. somes (6–8), have reduced the effort high-throughput method for the isola- © 2005 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0506279102 PNAS ͉ September 13, 2005 ͉ vol. 102 ͉ no. 37 ͉ 13001–13002 Downloaded by guest on September 28, 2021 cell surface that may play a trapped genes in public and private re- 20% of trappable genes that are not role in axon guidance. For this purpose, sources (available from the International currently represented in the public re- a secretory trap cassette (21) was Gene Trap Consortium, www.genetrap- source. Furthermore, a significant frac- flanked by the arms of homology to the .org) serves as a useful guide for identi- tion of trapped genes are represented by target gene. An astonishingly high fre- fying genes for targeted trapping. single events and thus the generation of quency of correctly targeted events was The work of Friedel et al. (5) comes additional mutant cell lines for these observed for 16 of the 24 loci, averaging at a propitious time and has important loci will be beneficial. Although most Ͼ50%. Importantly, Friedel et al. estab- strategic implications for international gene-trap mutations studied in mice ef- lish the threshold of expression above efforts aimed at generating a complete fectively mutate the endogenous gene at which targeted trapping is effective. the site of insertion, the generation of Semiquantitative RT-PCR was used to null alleles is not always guaranteed. measure steady-state mRNA levels of The efficiency Hypomorphic mutations (16) or inser- selected target genes relative to trans- tions that disrupt only a subset of ferrin receptor, a gene expressed at low of targeted trapping mRNA isoforms (22) are possible. The levels in ES cells. Efficient targeting was rapid generation of additional null al- observed for most genes expressed exceeds the current leles by targeted trapping will be valu- above 1% the level of transferrin recep- able in these cases. Finally, the mouse tor. A retrospective analysis showed that efficiency of random genetics community has expressed a 97% of genes previously trapped by ran- strong desire for reporter-tagged, condi- dom gene trapping are in fact expressed gene trapping. tional null alleles to enable temporal or above this threshold. Thus, genes acces- tissue-specific ablation of gene function sible to random trapping appear to be (1, 2). Targeted trapping can be readily equally good substrates for targeted collection of reporter-tagged null muta- adapted for conditional mutagenesis by, trapping. A survey of randomly selected tions in mice. The public gene-trap re- for example, incorporating conditional genes predicts that more than half of all sources contain mutations in Ϸ40% of gene-trap cassettes (23). genes are expressed above this threshold genes; however, the efficiency of trap- Gene targeting in ES cells will con- and, therefore, are likely to be accessi- ping new genes has dropped to Ϸ10% tinue to be the workhorse for the func- ble to random and targeted trapping. (one new gene is trapped for every 10 tional analysis of genes in mice for many This result agrees well with the fraction colonies isolated) and will continue to years to come. The detailed analysis of of genes that is thought to be accessible diminish well before saturation is genes will require all manner of alleles to gene trapping, estimated to be Ϸ60% achieved. The efficiency of targeted beyond the generation of null mutations. of all mouse genes (14). Therefore, most trapping exceeds the current efficiency Targeted trapping is a welcome addition of the genes accessible to targeted trap- of random gene trapping; therefore, tar- to the arsenal of molecular tools with ping are already represented in the ex- geted trapping could be put to good use which to address gene function in the isting gene-trap libraries, and a list of to systematically target the remaining mouse.

1. Austin, C. P., Battey, J. F., Bradley, A., Bucan, M., ics 73, 56–65. drich, L. V., Lu, X., Avery, B. J., et al. (2001) Nat. Capecchi, M., Collins, F. S., Dove, W. F., Duyk, 9. Valenzuela, D. M., Murphy, A. J., Frendewey, D., Genet. 28, 241–249. G., Dymecki, S., Eppig, J. T., et al. (2004) Nat. Gale, N. W., Economides, A. N., Auerbach, W., 17. Hansen, J., Floss, T., Van Sloun, P., Fuchtbauer, Genet. 36, 921–924. Poueymirou, W. T., Adams, N. C., Rojas, J., E. M., Vauti, F., Arnold, H. H., Schnutgen, F., 2. Auwerx, J., Avner, P., Baldock, R., Ballabio, A., Yasenchak, J., et al. (2003) Nat. Biotechnol. 21, Wurst, W., von Melchner, H. & Ruiz, P. (2003) Balling, R., Barbacid, M., Berns, A., Bradley, A., 652–659. Proc. Natl. Acad. Sci. USA 100, 9918–9922. Brown, S., Carmeliet, P., et al. (2004) Nat. Genet. 10. Gossler, A., Joyner, A. L., Rossant, J. & Skarnes, 18. Zambrowicz, B. P., Abuin, A., Ramirez-Solis, R., 36, 925–927. W. C. (1989) Science 244, 463–465. Richter, L. J., Piggott, J., BeltrandelRio, H., Bux- 3. van der Weyden, L., Adams, D. J. & Bradley, A. 11. Friedrich, G. & Soriano, P. (1991) Genes Dev. 5, ton, E. C., Edwards, J., Finch, R. A., Friddle, C. J., (2002) Physiol. Genomics 11, 133–164. 1513–1523. et al. (2003) Proc. Natl. Acad. Sci. USA 100, 4. Stanford, W. L., Cohn, J. B. & Cordes, S. P. (2001) 12. Skarnes, W. C., Auerbach, B. A. & Joyner, A. L. 14109–14114. Nat. Rev. Genet. 2, 756–768. (1992) Genes. Dev. 6, 903–918. 19. Schwartzberg, P. L., Goff, S. P. & Robertson, E. J. 5. Friedel, R. H., Plump, A., Lu, X., Spilker, K., 13. Zambrowicz, B. P., Friedrich, G. A., Buxton, E. C., (1989) Science 246, 799–803. Jolicoeur, C., Wong, K., Venkatesh, T. R., Yaron, Lilleberg, S. L., Person, C. & Sands, A. T. (1998) 20. Sedivy, J. M. & Sharp, P. A. (1989) Proc. Natl. A., Hynes, M., Chen, B., et al. (2005) Proc. Natl. Nature 392, 608–611. Acad. Sci. USA 86, 227–231. Acad. Sci. USA 102, 13188–13193. 14. Skarnes, W. C., von Melchner, H., Wurst, W., 21. Skarnes, W. C., Moss, J. E., Hurtley, S. M. & 6. Zhang, Y., Muyrers, J. P., Testa, G. & Stewart, Hicks, G., Nord, A. S., Cox, T., Young, S. G., Ruiz, Beddington, R. S. (1995) Proc. Natl. Acad. Sci. A. F. (2000) Nat. Biotechnol. 18, 1314–1317. P., Soriano, P., Tessier-Lavigne, M., et al. (2004) USA 92, 6592–6596. 7. Yu, D., Ellis, H. M., Lee, E. C., Jenkins, N. A., Nat. Genet. 36, 543–544. 22. Burgess, R. W., Skarnes, W. C. & Sanes, J. R. Copeland, N. G. & Court, D. L. (2000) Proc. Natl. 15. Leighton, P. A., Mitchell, K. J., Goodrich, L. V., (2000) J. Cell. Biol. 151, 41–52. Acad. Sci. USA 97, 5978–5983. Lu, X., Pinson, K., Scherz, P., Skarnes, W. C. & 23. Schnutgen, F., De-Zolt, S., Van Sloun, P., Hollatz, 8. Lee, E. C., Yu, D., Martinez de Velasco, J., Tessier-Lavigne, M. (2001) Nature 410, 174–179. M., Floss, T., Hansen, J., Altschmied, J., Seisen- Tessarollo, L., Swing, D. A., Court, D. L., 16. Mitchell, K. J., Pinson, K. I., Kelly, O. G., Brennan, berger, C., Ghyselinck, N. B., Ruiz, P., et al. (2005) Jenkins, N. A. & Copeland, N. G. (2001) Genom- J., Zupicich, J., Scherz, P., Leighton, P. A., Goo- Proc. Natl. Acad. Sci. USA 102, 7221–7226.

13002 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0506279102 Skarnes Downloaded by guest on September 28, 2021