Downloaded from genome.cshlp.org on September 27, 2021 - Published by Cold Spring Harbor Laboratory Press Perspective Hunting with Traps: Genome-Wide Strategies for Gene Discovery and Functional Analysis Kyle Durick, John Mendlein, and Kleanthis G. Xanthopoulos1 Aurora Biosciences Corporation, San Diego, California 92121 USA With sequence analysis of the human genome well underway, there is an increasingly urgent challenge to understand the fundamental function and interplay of genes that build and maintain an organism. Several approaches will be critical for interpreting gene function, including random cDNA sequencing, expression profiling in different tissues, genetic analysis of human or model organism phenotypes, and creation of transgenic or “knockout” animals. Traditional gene-trapping approaches, in which genes are randomly disrupted with DNA elements inserted throughout the genome, have been used to generate large numbers of mutant organisms for genetic analysis. Recent modifications of gene-trapping methods and their increased use in mammalian systems are likely to result in a wealth of new information on gene function. Various trapping strategies allow genes to be segregated based on criteria like the specific subcellular location of an encoded protein, the tissue expression profile, or responsiveness to specific stimuli. Genome-wide gene-trapping strategies, which integrate gene discovery and expression profiling, can be applied in a massively parallel format to produce living assays for drug discovery. Classical Genetics, Genomics, and Analysis through random sequencing of cDNA libraries derived of Gene Function from specific tissues or cell types (Okubo et al. 1992; The term “genomics” was proposed in 1986 by Thomas Fannon 1996; Okubo and Matsubara 1997). Analysis of Roderick (McKusick 1997a) to describe the study of a these expressed sequence tags (ESTs) can provide an genome by molecular means distinct from traditional “in-silico” gene expression landscape. Shifting towards genetic approaches. Genomics evolved from a much the gene products, large two-dimensional polyacryl- older word, genome (McKusick 1997a). A fusion of amide gels can be used to monitor the expression of at gene and chromosome, a genome is the complete col- least a subset of proteins (Geisow 1998; Persidis 1998). lection of genes possessed by an organism. Living crea- In addition, various methods are available for elucidat- tures result from the delicate interplay between a ing physical interactions between proteins in live cells “functional” genome and environmental factors. With (Fields and Song 1989; Miyawaki et al. 1997). complete sequence of many genomes, including bio- Classical and molecular genetics provide a variety logical workhorses like Escherichia coli (Blattner et al. of powerful tools useful for understanding gene func- 1997), Saccharomyces cerevisiae (Mewes et al. 1997), and tion and studying complex developmental events. Caenorhabditis elegans (The C. elegans Sequencing Con- Many structure–function relationships have been elu- sortium 1998), and with the Human Genome Project cidated using elegant genetic approaches and by clon- “ahead of schedule and under budget” (Collins 1995; ing disease associated genes. However, these strategies Collins et al. 1998), the scientific focus is shifting to- are too time consuming for efficient genome-wide in- wards development of methods that effectively use this vestigation. This review will focus on the use of ran- wealth of structural genomic information (Collins et dom gene-trapping techniques to gain insights into al. 1998). gene function. These methods are producing valuable The ultimate goal of the Human Genome Project is raw material for current and future bioinformatics da- to understand how the genome builds, maintains, and tabases that will catalog many biological processes. Ul- operates an organism. Multiple strategies at the DNA, timately, this information should help researchers pre- RNA, protein, cellular, and organism level will be key dict gene function in much the same way sequence in achieving this goal. Hybridization-based tech- databases are currently used in structural genomics ef- niques, like oligonucleotide chips (Fodor et al. 1991) or forts. cDNA arrays (Schena et al. 1995), can reveal the differ- ential expression profiles of numerous genes in re- Gene Capturing and Mutagenesis sponse to a stimulus (DeRisi et al. 1997; Wodicka et al. Conceptually, one straightforward way to discern the 1997). Gene expression profiles can also be elucidated function of a gene is to observe the effects of an im- paired or mutated gene on an organism over a desired 1Corresponding author. number of generations. Many gene mutations have E-MAIL [email protected]; FAX (619) 404-6719. consequently been implicated in a variety of human 9:1019–1025 ©1999 by Cold Spring Harbor Laboratory Press ISSN 1054-9803/99 $5.00; www.genome.org Genome Research 1019 www.genome.org Downloaded from genome.cshlp.org on September 27, 2021 - Published by Cold Spring Harbor Laboratory Press Durick et al. medical conditions. In addition, chromosomal mark- tern of the trapped gene can be visualized. In Dro- ers identified through arduous structural genomics ap- sophila, the P-transposon system has been used with proaches have led to remarkable successes in positional great success to create transformed animals carrying cloning of disease related genes. The important func- enhancer detectors (Rubin and Spradling 1982; O’Kane tional insights gained through time-consuming map- and Gehring 1987). Large numbers of enhancer trap based discovery of genes like the cystic fibrosis trans- lines have been established and evaluated using both membrane conductance regulator (Riordan et al. 1989; phenotypic and expression analysis (Bier et al. 1989; Rommens et al. 1989) or BRCA-l (Miki et al. 1994) have Spradling et al. 1995). Specific mutant lines can then resulted in the mapping of many phenotypic varia- be chosen for further study when there is a correlation tions to a specific gene. Much of the genetic basis of between expression and phenotype. disease work has culminated in the Online Mendelian Although enhancer traps are widely used, certain Inheritance in Man (OMIM) database provided on the gene traps were developed as an alternative to en- National Center for Biotechnology Information (NCBI) hancer traps to capture open reading frame informa- Web site (http://www.ncbi.nlm.nih.gov/Omim/), tion. The identification of target genes using enhancer which currently lists ∼500 human genes that have dis- trapping was sometimes problematic because the site ease-producing mutations (McKusick 1997b). of reporter insertion could be as much as 100 kb from Nonhuman organisms have been exploited to the target gene. This would require extensive charac- gather additional information about gene function. terization of the genomic insertion site to identify can- Each organism offers its own unique advantages and didate target genes. Gene trapping varies from en- drawbacks. Yeast, as the simplest eukaryote, is a logical hancer trapping in that, instead of using a minimal place to start when searching for basic understanding promoter, gene trap vectors provide specific sequences of cell biology. Studies with C. elegans, Drosophila, ze- that generate fusion RNA transcripts when inserted brafish, and other model systems have revealed the into a gene (See Fig. 1). This feature makes gene trap- functions of many genes during embryonic develop- ping (vs. enhancer trapping) especially advantageous ment or complex intercellular signaling. These organ- in mammalian cells that have complex genomic orga- isms, however, are very distantly related to humans; so nization, including large introns and small exons, be- mammalian systems are required to expand the knowl- cause the trapped gene can be identified by mRNA se- edge base and ultimately for pharmacological evalua- quence. tions. The combined use of embryonic stem (ES) cells It is believed that the use of random “knockout” with homologous recombination in mice has created a mutagenesis by gene trapping in mice will result in very useful system for functional genomic study, al- further enhancements for functional analysis of the lowing researchers to modify or eliminate any known genome. Typical vectors used for knockout gene trap- gene and analyze its function (Capecchi 1989; Joyner ping contain an acceptor site for RNA splicing followed 1993). by a reporter gene and then a transcript-terminating In contrast to the rational “one gene at a time” polyadenylation sequence (Brenner et al. 1989; Gossler approaches, enhancer-trapping methods evolved to et al. 1989; Friedrich and Soriano 1991; Skarnes et al. probe the entire genome simultaneously. Originally 1992). These gene-trapping vectors can be introduced described in bacteria, enhancer trapping was first dem- into the genome of murine ES cells by electroporation onstrated using bacteriophage transposable elements or replication-deficient self-inactivating retroviruses. to insert a reporter gene at scattered sites throughout Insertion of this vector into an intron results in pre- the E. coli genome (Casadaban and Cohen 1979; Bel- mature termination of the captured allele in which the lofatto et al. 1984). Chromosomal integration of a splice donor at the 3Ј end of an endogenous exon is transposable element tagged the integration site and “trapped” into splicing with the splice acceptor from often mutated the gene into which it inserted.