While it is important to know the sequence of a gene and its temporal-spatial pattern of expression, what’s really crucial is to know the functions of that gene during development. Recently developed techniques have enabled us to study gene function by moving certain genes into and out of embryonic cells.
Cloned pieces of DNA can be isolated, modified (if so desired), and inserted into cells by several means. One direct technique is microinjection, in which a solution containing the cloned gene is injected into the nucleus of a cell (Capecchi 1980). This technique is especially useful for inserting genes into newly fertilized eggs, since the haploid nuclei of the sperm and egg are relatively large (Figure 1). In transfection, DNA fragments may be incorporated directly into cells by incubating the cells in a solution designed to make them “drink” the new DNA in. The chances of a DNA fragment being incorporated into the chromosomes in this way are relatively small, however, so the DNA of interest is usually mixed with another gene, such as a gene encoding resistance to a particular antibiotic, that enables the rare cells that do incorporate the DNA to “identify themselves” by surviving under culture conditions that kill all the other cells (Perucho et al. 1980; Robins et al. 1981). A similar technique is electroporation, in which a high-voltage pulse “pushes” the DNA into the cells.
A more “natural” way of getting genes into cells is to insert the cloned gene into a transposable element or retroviral vector. These naturally occurring mobile regions of DNA can integrate themselves into the genome of an organism. Retroviruses are RNA-containing viruses. They enter a host cell, where they make a DNA copy of themselves (using their own virally encoded reverse transcriptase); the copy then becomes double-stranded and integrates itself into a host chromosome. The integration is accomplished by two identical sequences (long terminal repeats) at the ends of the retroviral DNA. Retroviral vectors can be made by removing the viral packaging genes (needed for the exit of viruses from the cell) from the center of a mouse retrovirus. This extraction creates a vacant site where other genes can be placed. By using the appropriate restriction enzymes researchers can excise a gene of interest (such as a gene isolated by PCR) and insert it into a retroviral vector. Retroviral vectors infect mouse cells with an efficiency approaching 100 percent.
Similarly, in Drosophila, new genes can be carried into the fly embryo via P elements. These DNA sequences are naturally occurring transposable elements that can integrate like viruses into any region of the Drosophila genome. These elements can be isolated and cloned genes can be inserted into the center of the isolated P element. When the recombined P element is injected into a Drosophila oocyte, it can integrate itself into the embryo’s DNA, providing the organism with the new gene (Spradling and Rubin 1982).
The techniques described above have been used to transfer genes into every cell of the mouse embryo. During early mouse development, there is a stage (the blastocyst) when only two cell types are present: the outer trophoblast cells, which will form the fetal portion of the placenta, and the inner cell mass (ICM), whose cells will give rise to the embryo itself. Separation of ICM cells can lead to twins (see Chapter 14), and if an ICM blastomere from one mouse is transferred into the embryo of a second mouse, that donor ICM cell can contribute to every organ of the host embryo.
Inner cell mass blastomeres can be isolated from an embryo and cultured in vitro; such cultured cells are called embryonic stem cells (ES cells). ES cells are almost totipotent, since each of them can contribute to all tissues except the trophoblast if injected into a host embryo (Gardner 1968; Moustafa and Brinster 1972). Moreover, once in culture, these cells can be treated as described in the preceding section so that they will incorporate new DNA. This added gene (the transgene) can come from any organism. A treated ES cell (the entire cell, not just the DNA) can then be injected into another early-stage embryo, and will integrate into this host. The result is a chimeric mouse* (Figure 2). Some of the chimera’s cells will be derived from the host’s own inner cell mass, but some portion of its cells will be derived from the treated embryonic stem cell. If the treated cells become part of the germ line of the mouse, some of its gametes will be derived from the donor cell. If such a chimeric mouse is mated with a wild-type mouse, some of its progeny will carry one copy of the inserted gene. When these heterozygous progeny are mated to one another, about 25 percent of the resulting offspring will be homozygous for (i.e., carry two copies of) the inserted gene in every cell of their bodies (Gossler et al. 1986). Thus a gene cloned from some other organism will be present in both copies of the chromosomes within these mouse genomes. Strains of such transgenic mice have been particularly useful in determining how genes are regulated during development.
The analysis of early mammalian embryos has long been hindered by our inability to breed and select animals with mutations that affect early embryonic development. This block has been circumvented by the techniques of gene targeting (or, as it is sometimes called, gene knockout). These techniques are similar to those that generate transgenic mice, but instead of adding genes, gene targeting replaces wild-type alleles with mutant ones. As an example, we will look at a knockout of the gene for bone morphogenetic protein 7 (BMP7). Bone morphogenetic proteins are involved in numerous developmental interactions whereby one set of cells interacts with other neighboring cells to alter their properties. BMP7 has been implicated as a protein that prevents cell death and promotes cell division in several developing organs.
Dudley and his colleagues (1995) used gene targeting to find the function of BMP7 in the development of the mouse. First, they isolated the Bmp7 gene, cut it at one site with a restriction enzyme, and inserted a bacterial gene for neomycin resistance into that site (Figure 3). In other words, they mutated the Bmp7 gene by inserting into it a large piece of foreign DNA that destroyed the ability of the gene’s product (BMP7 protein) to function. These mutant Bmp7 genes were electroporated into embryonic stem cells that were naturally sensitive to neomycin. Once inside the nucleus of an ES cell, the mutated Bmp7 gene may replace a normal allele of Bmp7 by a process called homologous recombination. In this process, the enzymes involved in DNA repair and replication incorporate the mutant gene in the place of the normal copy. It’s a rare event, but such cells can be selected by growing the ES cells in neomycin. Most of the cells are killed by the drug, but the ones that have acquired resistance from the incorporated gene survive. The resulting cells have one normal Bmp7 gene and one mutated Bmp7 gene. These heterozygous ES cells were then microinjected into mouse blastocysts, where they were integrated into the cells of the embryo. The resulting chimeric mice had wild-type cells from the host embryo and heterozygous Bmp7-containing cells from the donor ES cells. The chimeras were mated to wild-type mice, producing progeny that were heterozygous for Bmp7. These heterozygous mice were then bred with each other, and about 25 percent of their progeny carried two copies of the mutated Bmp7 gene. These homozygous mutants lacked eyes and kidneys (Figure 4). In the absence of functional BMP7 protein, it appears that many of the cells that normally form these two organs stop dividing and die. In this way, gene targeting can be used to analyze the roles of particular genes during mammalian development.