Biotechnology News
- Newsletter - Volume 47
- Newsletter - Volume 46
- Newsletter - Volume 45

Other Links
- School of Biosciences

Understanding gene function and genetically based disease using mouse models

Jon Frampton

Institute for Biomedical Research, Birmingham University Medical School

From genome sequence to medicine

The human genome sequence has already provided answers to many questions. However, as with any achievement in science, the number of new questions generated has far outstripped those that have been solved. What will knowledge of the human genome mean for medical research? Inherited and acquired DNA sequence differences are the root cause of many human diseases. Loss or mutation of an essential protein can directly cause disease as seen with the absence of globin gene expression in thalassaemia or aberrant function of mutated CTFR protein in cystic fibrosis. In contrast, disease susceptibility manifested as an increased life-long risk (eg cardiovascular disease) probably results from a genetic “profile” involving more subtle differences in several, perhaps many, genes. Hence, although it is relatively straightforward to identify the genetic change underlying diseases in the former category it is a considerable challenge to elucidate the sequences specifying complex genetic diseases.
         Genetic mapping and comparative sequencing projects can identify associations between DNA sequence and a disease or human trait, but it remains necessary to test this link. Moreover, once DNA sequence differences in a gene or genes have been identified as the cause of disease how can we move forward towards the development of therapies? The incredible advances over the last 20 years in our ability to manipulate the genome of inbred strains of laboratory mice has given us the opportunity to address these issues. My aim in this article, taking examples from my own research, is to show how mouse genetics can form the cornerstone for investigative and translational biomedical research during the coming decade, and how researchers in Birmingham University will be at the forefront of this international effort.

The basis of genome alteration - the embryonic stem cell

Mouse genetic technology enables the removal or replacement of specific gene sequences, the consequences of which can then be assessed by studying the development of the mutated animals from the egg through to the adult and by careful analysis of biological systems and physiological processes. Key to the whole technology is the embryonic stem (ES) cell. These cells, which can be derived from the early pre-implantation embryo (“blastocyst”), can be grown indefinitely in culture while retaining the ability to give rise to a mouse if injected back into a blastocyst (the “host”). Once re-implanted into the uterus of a foster mother, the injected (“donor”) ES cells compete with cells in the host blastocyst to form the developing embryo and ultimately lead to a chimaeric mouse formed from two genetically distinct cell types. If the germ cells (sperm or eggs) of the chimaera also contain cells derived from the donor ES cells then some progeny resulting from mating will have one set of chromosomes derived completely from the donor, thereby establishing a “line” (Figure 1). The other feature of ES cells that is crucial to their use in mouse genetics is that an introduced segment of mouse genome can recombine precisely with the corresponding sequences in the ES cell genome (“homologous recombination”). If the introduced segment contains an alteration (eg a deletion, insertion or small sequence change), then this will become a part of the ES cell genome and therefore ultimately can be included in the genome of a mouse (Figure 2).

“Simple” genetic alterations

The simplest, and so far most widely applied, use of homologous recombination in ES cells has been the generation of mice that no longer express a particular gene. This is commonly termed a “knockout” or null allele of a gene. Usually, all or part of the protein coding sequences is deleted or an insertion is made into an exon encoding an essential domain. The modified segment of genomic DNA also incorporates a positive selection gene (usually a bacterial antibiotic resistance gene). The so-called “targeting construct” is introduced into ES cells by subjecting a cell suspension in a solution of the DNA to a short electric pulse. ES cells that successfully exchange the targeting sequences by homologous recombination for one of its two corresponding genomic sequences are able to grow in the presence of the selective antibiotic and can expand as a clone. After extensive characterisation to ensure that the introduced sequences have correctly replaced the endogenous gene the ES cell clone is used to create a mouse line. The consequences of the loss of function of the gene are then usually assessed by breeding mice so that both copies are the knockout allele (homozygous null, or “nullizygous”).
         Although the process is now feasible if appropriate cell culture and animal facilities are available, gene knockouts cannot be regarded as a standard laboratory technique and a time scale of about a year to obtain a knockout mouse is probably realistic. The making of simple gene knockouts by individual laboratories may soon be a thing of the past as a number of companies develop high-throughput technologies for ES cell-targeting aiming to provide “off the shelf” knockouts of every single gene (Valenzuela et al, 2003).
         The effect of homozygosity for the knockout allele of a gene depends on the encoded protein’s role. If related proteins can perform a similar function then the absence of one of the corresponding genes may have little or no effect. This often means that several members of a gene “family” have to be deleted and be brought together in one mouse by lengthy cross-breeding before a phenotype is apparent. Even if a gene is not part of a family, its deletion may have subtle effects that can only be teased out by careful observation or by subjecting the animal to specific conditions. The essential function of some gene products means that homozygous knockout animals may die at some point in their development. Such a lethal effect is clearly informative, but may reflect the function of the protein in only a subset of the tissues in which it plays a role. Moreover, the death of a knockout animal at some stage during its development precludes analysis of the protein’s role in the adult.
         My own research on proteins that regulate gene expression during the production of blood cells and the vascular system in the developing foetus and the adult illustrates the sorts of information that can be obtained from gene knockouts. Our main focus is on the c-Myb protein, which controls the expression of genes necessary for the expansion and differentiation of the eight blood cell lineages and components of the vasculature. The association of c-Myb with the development of leukaemia and atherosclerotic plaques in arterial walls has raised considerable interest in its possibilities as a therapeutic target (Frampton, 2004). The c-myb gene was one of the first genes to be knocked out and showed through the death of embryos from anaemia that c-Myb is crucial for development of blood cells (Mucenski et al, 1991). Subsequent analysis by us revealed more about the defects caused by the absence of c-Myb in knockout embryos (Sumner et al, 2000). However, further understanding of the role of c-Myb in blood cells, and the chance to investigate it’s function in other tissues such as the vasculature, had to await the development of more sophisticated approaches, as I will describe below.

More complicated tricks of the trade - getting beyond a lethal phenotype

A number of elegant genetic methods are available to meet the challenge of investigating genes such as c-myb that have a lethal phenotype when knocked out. I will concentrate on the most widely used technology, namely conditional gene deletion. The basic principle is simple - a gene is engineered by homologous recombination in ES cells so that the whole gene or an exon encoding a crucial protein domain is flanked by recognition sites for a recombinase enzyme that can delete the intervening sequences. Gene knockout is restricted by expressing the recombinase in specific tissues or at a particular time. Several recombinase systems have been characterised from bacteria or yeast, but by far the most widely used at the moment employs the P1 bacteriophage phage-derived Cre recombinase and its 32 base recognition element loxP (Figure 3A). The Cre-loxP approach has been applied to a wide range of genes (Kwan, 2002), including c-myb in our laboratory (Emambokus et al, 2003). Although the principle of the method is simple, its application is not trivial. Generation of the targeting construct is much more complicated than for a simple knockout. Since the aim of such targeting is to generate an allele that behaves in the same way as the wild type until Cre recombinase brings about its deletion, provision is included for removal of the selection cassette using a second recombinase, Flp (Figure 3B). An important aspect of the technology is to express Cre recombinase where and when you want at sufficient levels to bring about efficient recombination. This involves crossing mice containing the loxP modification to strains expressing Cre in a particular cell type. For example, we use a strain containing a Cre transgene driven by the promoter of the smooth muscle myosin heavy chain gene to bring about recombination of the loxP-flanked c-myb gene in vascular smooth muscle cells. Alternatively, the recombinase can have the potential to be expressed in all tissues, but only when an inducer is provided. An example of the latter, which is frequently used in our research, makes use of a transgene containing an interferon-inducible promoter to drive Cre expression.

Mouse genetics is more than gene deletion

Although deletion is a very useful way of investigating the importance of a gene, the majority of diseases with a genetic basis involve small sequence changes. The same ES cell-based approaches can be used to create mouse lines to model such mutations (so-called “knockins”, Figure 3C). Many genetic diseases can arise from any one of several possible mutations in a particular gene (eg in the CTFR gene in cystic fibrosis). The prospect of mutating ES cells repeatedly with specific targeting constructs is a daunting one. However, recent developments, currently on trial in our laboratory, make use of the recombinase systems not to delete genes but rather to replace a given coding sequence in ES cells with multiple alternatives (Figure 3D). This technique, known as “recombination mediated cassette exchange”, can still require the costly and time-consuming derivation of mice from the mutated ES cells. Nevertheless, there is some prospect that this may be streamlined if the cassette exchange can be performed directly in oocytes (Lauth et al, 2000) or if mice can be produced by fusion of two different cells, rather than blastocyst injection (Nagy et al, 1990).
         From the point of view of our work on c-Myb, we are eager to create multiple single amino acid changes in protein domains implicated in interaction with co-operating factors. Interest in this stems from observations that reducing either c-Myb protein levels using a weakly expressed (“knockdown”) allele of c-myb (Emambokus et al, 2003) or c-Myb function by inhibiting its association with the protein CBP/p300 leads to dramatically elevated blood platelet levels. It is possible that this study could shed light on human diseases such as essential thrombocytosis that exhibit elevated platelet numbers.

Conclusions and the future of mouse genetics research in Birmingham University

Although my group’s work is focused on understanding the normal and disease-associated function of c-Myb and other proteins involved in the regulation of blood and vascular cells, we want to push forward new technological developments that will increase our efficiency and the sophistication of our mouse models. To that end, and in collaboration with Dr Martyn Bell and Professor Eric Jenkinson in the Institute for Biomedical Research, and Dr Heiner Schrewe and Professor John Heath in Biosciences, we are endeavouring to establish state-of-the art facilities for the mutation of ES cells and the subsequent generation, maintenance and analysis of mouse lines. Hopefully, this will result in cutting edge mouse genetics in Birmingham University that will encourage others in the area of biomedicine to consider this avenue for their research.

Top