The Use of Animal Models in Studying Genetic Disease: Transgenesis and Induced Mutation

As the name might imply, indirect approaches attempt to randomly make mutations in animal models' genomes. Then, the animals are screened in an attempt to determine which ones show phenotypes that are similar to human diseases. Thus, instead of being driven by the disease mutation, these methods are based on screening the phenotypes.

Two of the most effective ways to generate mutations are by exposing organisms to X-rays or to the chemical N-ethyl-N-nitrosourea (ENU). X-rays often cause large deletion and translocation mutations that involve multiple genes (Bedell et al., 1997a), whereas ENU treatment is linked to mutations within single genes, such as point mutations (Hardouin & Nagy, 2000). ENU can produce mutations with many different types of effects, such as loss and gain of function (Rosenthal & Brown, 2007), and it is frequently employed in screens in model organisms such as zebra fish. These types of models are particularly useful for identifying new genes and pathways that contribute to disease.

As opposed to the use of X-rays and ENU, transgenesis is a directed approach. Transgenic animals are generated by adding foreign genetic information to the nucleus of embryonic cells, thereby inhibiting gene expression. This can be achieved by either injecting the foreign DNA directly into the embryo or by using a retroviral vector to insert the transgene into an organism's DNA. The first mouse gene transfers were performed in 1980 (Hardouin & Nagy, 2000); however, at that time, the methods for transgenesis were not optimal. For instance, the foreign DNA was incorporated into only a small percentage of embryos and was inconsistently passed to the next generation. Also, small transgenes were inserted into random sites in the genome, and depending on their location, they weren't always expressed. More recently, scientists have developed a way to increase the size of the DNA fragments used in transgenesis by cloning them in yeast or bacterial artificial chromosomes (YACs or BACs, respectively). These larger transgenes are more likely to contain regulatory sequences necessary for normal gene expression and are usually more comparable to the endogenous gene (Bedell et al., 1997a). As a result, the use of transgenic mice has dramatically increased in the past two decades, and this type of animal model has contributed greatly to our knowledge of disease development.

Both knock-out and knock-in models are ways to target a mutation to a specific gene locus. These methods are particularly useful if a single gene is shown to be the primary cause of a disease. Knock-out mice carry a gene that has been inactivated, which creates less expression and loss of function; knock-in mice are produced by inserting a transgene into an exact location where it is overexpressed. Over the years, more than 3,000 genes have been knocked out of mice, and most of these genes have been related to disease (Hardouin & Nagy, 2000).

Both knock-out and knock-in animals are created in the same way: a specific mutation is inserted into the endogenous gene, and then it is conveyed to the next generation through breeding. The use of embryonic stem (ES) cells is required for this technology. This is because ES cells can contribute to all cell lineages when injected into blastocysts, and they can be genetically modified and selected for the desired gene changes. Homologous recombination creates the mutations; this is a process that physically rearranges two strands of DNA for the exchange of genetic material. Many types of mutations can be introduced into a model gene in this way, including null or point mutations and complex chromosomal rearrangements such as large deletions, translocations, or inversions (Bedell et al., 1997a). Many knock-out and knock-in mice have similar, if not identical, phenotypes to human patients and are therefore good models for human disease.

One drawback to using transgenic, knock-in, and knock-out mice to study human diseases is that many disorders occur late in life, and when genes are altered to model such diseases, the mutations can profoundly affect development and cause early death. These effects would preclude using animal models to study adult diseases. Thankfully, new technology has made it possible to generate mutations in specific tissues and at different stages of development, including adulthood. To do this, mice with two different types of genetic alterations are needed: one that contains a conditional vector, which is like an "on switch" for the mutation, and one that contains specific sites (called loxP) inserted on either side of a whole gene, or part of a gene, that encodes a certain component of a protein that will be deleted (Bedell et al., 1997a). A conditional vector for the gene is made by inserting recognition sequences for the bacterial Cre recombinase (loxP sites) using homologous recombination in ES cells. The vector contains a drug-resistant marker gene that allows only the targeted ES cells to survive when exposed to the drug. Thus, the mutant ES cells can be selected and injected into the host mouse embryo, which is implanted into a foster mother. The resulting offspring are chimeras and have multiple populations of genetically distinct cells. Chimeric offspring are then crossed, and the resulting generation of offspring has the recombinase effector gene. The mice containing the Cre recombinase under the control of tissue-specific or inducible regulatory elements are crossed to the mice with the desired loxP sites. When Cre is expressed, recombination occurs at the loxP sites, which delete the intervening sequences, and the resulting mutation is induced in specific regions and times. These conditional mutant models are becoming increasingly popular, and international initiatives have been created to accommodate their demand (Rosenthal & Brown, 2007).

The aforementioned advances in ES cells and Cre/loxP conditional mutations have helped pave the way for the creation of models for complex human diseases involving chromosomal rearrangements. Mouse models of these disorders can be created using indirect approaches, such as radiation, but their usefulness is restricted because pathological endpoints are unpredictable and undefined (Yu & Bradley, 2001). Using the Cre/loxP recombination system overcomes these setbacks by allowing site-specific mutations necessary to produce accurate models of defects caused by human chromosomal rearrangements. These mutations can include chromosome deletions, duplications, inversions, and translocations, as well as nested chromosome deletions.