MCDB 4620            DEVELOPMENTAL BIOLOGY LECTURW NOTES            11/5/02
Lecture 16 (Jones)      Mouse developmental genetics and transgenic technology

Reading: Reading: Gilbert 4: 87-89, 97-102 6 : 164

The development of mammals is difficult to study in many regards. Among mammals, the mouse has been the most favored organism because of its relatively small size (reducing housing costs) and relatively short generation time. However, when compared to C. elegans or Drosophila, mice have a considerably longer generation time, are expensive to maintain, and are less accessible to experimental manipulation and observation as embryos because they develop in utero. Nevertheless, mouse development has become an intensive area of study because of its obvious relevance to human development and disease. Techniques for introducing cloned genes into the germ line to make transgenic animals have greatly increased the potential of mouse developmental genetics. Similar techniques are being applied to other animals and plants with practical benefits for agriculture.

Mouse developmental genetics

Practical considerations. Laboratory mice have a generation time of about 8 weeks and a genome size about the same as humans, 3000 Mb (about 30 times C. elegans or 15 times Drosophila), carried on 20 chromosomes. Mice can be mutagenized with X-rays or with chemical mutagens such as ethylnitroso-urea (ENU), which tends to induce single base changes and is administered by peritoneal injection into males. Many induced or spontaneous mutations have been mapped and studied over the past 50 years, including some that result in developmental defects. However, because of the expense of maintaining large populations and mutant stocks, saturation screening for a phenotype of interest as with Drosophila or C. elegans has been only infrequently attempted. Moreover, cloning of mutationally identified genes is more difficult because of the much larger genome size, although recent achievement of a high resolution genetic map (having molecular markers approximately every 400 kb) is making this easier. The progress of the mouse genome project, and the similarity of mice to humans have within the last couple of years encouraged new efforts to begin large scale saturation mutagenesis projects using the mouse.

Genetic control of stem cell development. An example of successful forward genetic analysis from several years ago is the discovery of two genes involved in control of stem cell proliferation, defined by dominant mutations at the Steel (Sl) and White spotting (W) loci. Mutations at both loci result in reduced numbers and/or defective migration of melanocytes (neural crest-derived skin pigment cells), hematopoietic stem cells, and primordial germ cells, which in turn lead to abnormal coloring, anemia, and sterility. Similar mutations have been observed in humans. Phenotypic similarity of Sl and W mutations suggested they might act in the same pathway. In transplant experiments using wild-type or mutant bone marrow cells and mutant or wild-type recipient mice, respectively, it was shown that W defects were stemcell autonomous, whereas Sl defects were not. With difficulty and some good luck (this was early days of cloning), it proved possible to clone both loci. The W locus encodes a receptor RTK that had been previously identified as an oncogene (c-kit). The Sl locus was shown to encode a novel glycoprotein growth factor, which binds to the W receptor. Sl is expressed in the embryonic microenvironments where the three affected stem cells migrate and proliferate: namely, in the ectoderm and dorsal somite regions (melanocytes); in the yolk sac, embryonic liver, and developing bone marrow (hematopoietic stem cells); and along the hindgut, mesentery, and genital ridge (primordial germ cells).

Early experiments on manipulation of mouse embryos led to another technique by which mosaic or chimeric mice can be created, for use in answering questions about autonomy or non-autonomy like the one above. Recall that cells in the early mouse blastula are totipotent, and can develop into either trophoblast or inner cell mass (ICM – the embryo proper). If the zona pellucida is removed from each of two genetically different blastula stage embryos (before compaction) and the blastulae are pushed together in culture, the resulting fused blastula can be reimplanted into a foster mother and will develop into a chimeric mouse, with some tissues derived from one blastula (set of parents) and the rest derived from the other blastula (set of parents). Such mice are called tetraparental chimeras; they generally have equal numbers of cells derived from each of the two blastulae. Other experiments showed that unequal chimeras could be made by microinjection of dispersed embryonic cells in culture into a blastocyst, where they would become part of the ICM and contribute to tissues of the embryo and the resulting mouse. This technique became important later for the process of targeted mutagenesis we will consider in the next lecture. Subsequent work led to techniques for manipulating one-celled embryos and oocytes by microinjection. Replacement of pronuclei, to make embryos with two male or two female pronuclei, revealed the phenomenon of imprinting. Transplantation of somatic nuclei into enucleated oocytes eventually led to mammalian cloning. And microinjection of DNA into one pronucleus provided the first reliable method for making transgenic mice.
 
Transgenic mice

Introduction of cloned genes. The mouse became a much more useful organism for developmental geneticists with the advent of techniques for introducing cloned DNA sequences into the germ line. Viral genes were the first to be introduced, by microinjection of SV40 or infectious retroviruses into early embryos. Genes were also introduced by transfection of DNA into teratocarcinoma (totipotent embryonic tumor) cells followed by injection of selected cells into the mouse blastocyst, and their incorporation into the embryo. Alternatively, nuclei from such cells were introduced into fertilized eggs from which the pronuclei had been removed. However, the most successful approach turned out to be direct microinjection of DNA into pronuclei of fertilized eggs.

Integration and expression of injected DNA. DNA microinjected at the one-cell stage usually integrates prior to first cleavage at one apparently random chromosomal site, so that all cells of the resulting animal carry the integrated sequences, which are generally present in multiple copies arranged in tandem head-to-tail arrays. Integrated genes are generally expressed, although the level tends to vary from one transgenic mouse to another, presumably due to position effects of the integration site (What are position effects?). Intact genes are often expressed with normal temporal and tissue specificity. This has allowed systematic modification and reintroduction of cloned genes to identify sequences involved in developmental regulation.

Artificial regulation of expression. Mixing and matching of genes and regulatory elements allows engineering of cloned genes that will be expressed in new ways. Thus, dominant (or dominant negative) mutations can be created, and the phenotype of the resulting mice studied as a means of analyzing gene function in vivo in a mammalian system. One promoter has been that of the gene for metallothionein (MT), an inducible protein that binds strongly to and thereby protects the animal against heavy metal ions; it is synthesized at high levels in liver in response to heavy metals in the diet. A transgene fused to an MT promoter can be turned on by feeding low levels of a heavy metal to the mouse. This approach was used to express a MT-growth-hormone fusion gene, resulting in synthesis of high levels of growth hormone in the liver, bypassing the normal feedback controls on growth factor levels, and resulting in mice about twice the normal size. More recently, various tetracycline-inducible promoters have been developed, with the goal of being able to regulate gene expression rapidly and specifically in the mouse.

Insertional mutagenesis. Integration of injected DNA provides the opportunity for insertional mutagenesis of endogenous genes, which is the process became "tagged": for cloning. This technique is useful for identifying and isolating developmentally interesting genes, identified by a mutant phenotype (e.g.the reeler gene, in which an insertion mutation causes defective cell migration in the nervous system), or by an interesting expression pattern (e.g. enhancer and promoter traps).

Transgenic rescue. Transgenic technology offers the possibility of "gene therapy" for a defective gene, by what in worms or flies would be called transformation rescue. This goal was achieved in the late 1980s in mice carrying the shiverer mutation, which results in defective myelin basic protein (MBP) a component of the myelin sheath of neurons. Mice homozygous for the shiverer allele mld show decreased myelination in the CNS, tremors, and convulsions of increasing severity leading to early death. Introduction of an MBP transgene into mld embryos partially or completely cured these defects, in animals developing from transformed embryos as well as in their progeny.

Note that this procedure is very different from the "gene therapy" that is now being tested on humans for the treatment of a few heritable diseases. No one so far is attempting or seriously proposing to introduce transgenes into the human germ line, as was done with mice in the example above. Gene therapy in humans is intended to correct a gene defect only somatically, in the affected tissues of an afflicted individual, by introducing a normal cloned gene that will be expressed locally. Suitable vectors and methods of delivery are still being developed, but this approach appears promising for certain genetic diseases, as we will discuss further in a later lecture.

Agricultural applications

Some success has been achieved in producing transgenic animals and plants with new desirable traits by introducing cloned genes into the germ line. Although this technology is controversial in some circles, it should be remembered that genetic engineering of agriculturally important species has been going on for thousands of years in the form of selective breeding; and that some of the resulting species (for example, maize and cattle) are very different from the most closely related "natural" organisms. The newer molecular methods are simply faster and more versatile. Other "unnatural" methods for enhancing the modification of animal species by ordinary breeding have become widely accepted. Artificial insemination and more recently embryo transfer technology allow the widespread dissemination of desirable genomes; these techniques can be further enhanced by embryo splitting and nuclear transplantation procedures, or using the newer cloning procedures, to produce several identical individuals from a single animal. Some of these techniques are in widespread use in the livestock industry, and the newer techniques are likely to see increased use. Note that these practices have the undesirable consequence of decreasing variation in the gene pool.

Several transgenic animals and plants with desirable new traits are now being commercialized. As examples, it has been possible to introduce insect- and herbicide-resistance into cotton and maize, to create tomatoes which soften more slowly allowing them to be vine-ripened longer (a use of anti-sense technology!), to modify the oil content of canola through expression of lipid modifying enzymes in the plant, and to create goats which secrete recombinant proteins into their milk to provide a convenient source for large-scale purification. For some plant applications see

http://www.colostate.edu/programs/lifesciences/TransgenicCrops/

Reverse Genetics
Developmental  genetics ("forward genetics") typically proceeds by isolating a mutant with an interesting phenotype, mapping the mutation, then cloning and molecularly characterizing the gene and its products. The goal of reverse genetics is to do the opposite: starting with a cloned gene of interest, to somehow introduce mutations into the corresponding gene in the organism, and then analyze the resulting phenotype. This targeted mutagenesis can be done in theory by altering the cloned gene to make it defective, and then causing it to replace the normal resident gene by a homologous recombination (HR) event.

The problem of targeted mutagenesis by homologous recombination (HR)
DNA introduced into mammalian cells generally integrates randomly into chromosomal locations where nicks happen to be present. If HR occurs, it is rare and must be screened or selected for, so that introduction of sequences by HR cannot be done directly with embryos. Instead, cultured cells must be used, selected or screened for desired integrants using suitable markers, and then introduced into embryos in such a way that they will be incorporated into the germ line.

This can be done with embryonic stem (ES) cells. These pluripotent cells can be cultured from inner cell mass of early mouse embryos. Importantly, ES cells can give rise to essentially all tissues of the animal, including germ cells, allowing propagation of alterations of the ES cell genome through the germ line.

Does homologous recombination of transgenes occur in ES cells?
To answer this question, Mario Capecchi and colleagues at Univ. of Utah engineered a cloned hprt gene (which makes cells sensitive to thioguanine, 6-TG) disrupted by a bacterial neoR gene (which makes cells resistant to neomycin or the related drug G418). If this construct integrates into the genome at any location, the ES cell will become resistant to G418; if it integrates into the hprt gene (on the X chromosome) by HR, it will disrupt this gene and the cell will also become 6-TG-resistant (incorporation of 6-TG into DNA through activity of Hgprt enzyme kills cells). He found that HR does occur, but only at 0.1% the frequency of random integration. He then tested a few other engineered cloned genes similarly and demonstrated that HR was occurring with similar frequencies, assayed using PCR with appropriate primers (How?). So it should work; but how to increase the frequency, or identify these rare cells in which HR occurs?

Isogenic DNA and homologous recombination frequency
Subsequently, it has been found that a critical parameter in determining the HR efficiency is the degree of identity between the ES cell genome and homologous sequences in the targeting construct; only a few mismatches due to genetic polymorphisms  between different strains of mice can greatly lower the frequency of HR. Conversely, if both the ES cells and the construct used are derived from the same strain, it is not unusual for 1/20 of the G418-resistant ES clones from a transfection to be homologous recombinants! Consequently, in current practice it is often sufficient to use only positive selection (for G418 resistance) and screen the resulting colonies by PCR analysis or Southern blot analysis.

Making mutant mice
1- ES cells (typically homozygous for a dominant coat color marker allele) are cultured, transfected, selected with appropriate drugs (e.g. G418 in the case of the neo resistance gene), screened by PCR and Southern blot analysis to identify clonal cell lines that contain the targeted insertion (what will be the genotype of these cells?).

2- These cells are then introduced into mouse blastocysts by injection and implanted in pseudopregnant mothers. If mice having the color expected for the ES cells are born (recall a homozygous dominant coat color marker was present in the ES cell genome), they are checked for the mutation by DNA analysis and heterozygotes are intercrossed.

3- Some of the resulting progeny will be homozygous for the mutation and can be examined for a mutant phenotype.  If the initial transgenic construct was engineered to contain an inactive form or deletion of the gene (i.e. a null allele), the resulting homozygous transgenic animal is called a "knockout" mutant.

Related techniques harnessing the power of Homologous Recombination
"Knock-ins."  Using the HR technology, any sequence can be introduced into a genetic locus, including variants of the original sequence. The term "knock-in" refers to replacement of a normal gene not with a deleted or disrupted nonfunctional gene, as in a knock-out, but with an altered  form of the original, or with a different gene.  The neoR gene can be placed in an intron, downstream, or removed after isolation of ES cells having the targeted mutation so that it does not interfere with subsequent function of the knocked-in gene.

For example, we discussed in the last lecture how the results with reporter transgenes introduced randomly by embryonic injection are often confused by position effects, or by failure to include all regulatory elements in the reporter construct. A much more reliable method for determining expression patterns is to create a hybrid gene by integrating DNA encoding a reporter, such as lacZ or GFP, into the genetic locus, where it will be subject to normal gene controls.

Alternatively, instead of a reporter gene, any other gene can be expressed in the pattern of a targeted gene by integrating a cDNA into the locus. One powerful example is the use of knock-ins to test for functional homologies (or differences) between related genes. For example, it has been shown that the protein coding sequences of myogenin and Myf-5 (both basic HLH transcription factors in the same gene family) are functionally equivalent. If the coding sequences at the myogenin locus are replaced by those of Myf-5 in a mouse's genome, myogenesis is normal. A similar experiment was done with the mammalian homologs of Drosophila engrailed, En-1 and En-2 (these are both homeodomain transcription factors). It was found that knocking En-1 into the En-2 locus gave a normal mouse (En-1 null mutants die at birth with a large portion of their mid- and hindbrain missing). Think about it - is this surprising? What does it suggest about multigene families of developmental regulators in vertebrates?

Finally, knock-ins can be used to introduce essentially any desired mutation into the mouse. This application has been particularly useful in creating mouse models for human genetic diseases, where the molecular lesion in a human disease allele is known. For example, a point mutation in a gene known to cause Alzheimer's, or any other disease, can be introduced into the mosue, creating an animal model for the disease. This illustrates the remarkable flexibility of mouse reverse genetics with present state-of-the-art technology: literally any desired mutation can be generated!

"Cre-lox"
    If a null mutation causes embryonic lethality, it is not possible to directly observe the consequences of that mutation in an adult animal. If a gene is expressed in multiple tissues or at multiple times during development, it will often be desirable to create mosaic mutant animals or conditional mutants in order to ask where and when gene function is required. This can now be done in theory, and increasingly in practice, using the Cre-lox system, named for a bacteriophage site-specific recombinase (Cre) and the 34-bp tandem repeat sequences (lox) that it recognizes. Cre-mediated recombination can excise the DNA between two lox sequences. In mice, using HR, an exon of the gene to be analyzed can be replaced by a modified exon flanked by two lox sequences. Because there is no Cre recombinase activity in wild-type mice, this exon will be deleted and the gene thereby inactivated only in cells in which a Cre transgene is functional. By introducing Cre under the control of a tissue-specific promoter, it is possible to inactivate the target gene only in that tissue, creating a specific mosaic animal, and observe the consequences. Similar experiments can be done using a stage-specific promoter for Cre.

This potentially very powerful technique can also be used to investigate the functional role of individual RE's (by deleting them in a particular tissue), to selectively turn on a target gene by deleting a negatively acting RE (silencer element), or simply to "clean up" a transgenic organism by removing marker genes (e.g. neoR) that were introduced during its construction.

Other applications of mouse HR technology
This technique is leading to widespread application of the paradigm whereby a novel developmentally interesting gene is identified mutationally by forward genetics in C. elegans or Drosophila, and a mouse homologue is then found and deleted by HR reverse genetics to ascertain the gene's function in a mammal.

The technique will also likely be useful as an approach to somatic gene replacement therapy in humans, for conditions that can be corrected by isolating stem cells (e.g. blood stem cells to repair a globin defect) from a patient, engineering them in culture using a knock-in strategy to replace the defective (globin) gene with a normal one, and then re-introducing them into the appropriate tissue (bone marrow in this case).

In theory, replacement of germ-line genes could be accomplished in a similar manner using human ES cells, although if you think about how this might be done it would be much more difficult. This is probably an issue that society will have to deal with eventually.