ILAR Journal Online, Volume 38(1) 1997: Unusual Mammalian Models

Online Issues

<< All Back-issues

<< This Issue's Table of Contents

ILAR Journal V38(1) 1997
Unusual Mammalian Models

Transgenic Technology and Laboratory Animal Science
Jon W. Gordon

Jon W. Gordon, M.D., Ph.D., is Mathers professor of Geriatrics and Adult Development at the Mount Sinai School of Medicine in New York, New York.

INTRODUCTION

Transgenic technology allows the introduction of new and functional genetic material into the germ line. The first advanced eukaryotes to prove amenable to such genetic modifications were mice (Gordon and others 1980; Gordon and Ruddle 1981); subsequently, a variety of species including fish and fruit flies have been subject to transgenic manipulation. Germline gene transfer has enormously increased the rate of progress in understanding mammalian development. As a consequence of this improved knowledge, important insights into many disease processes have been gained, and new strategies for genetic engineering of livestock have been devised. These many advances have been comprehensively reviewed elsewhere (Palmiter and Brinster 1986; Gordon 1989). A related publication has also previously appeared in this ILAR News (Gordon 1988).

In this paper transgenic technology will be discussed with the issues of laboratory animal sciences as a reference point. The extant approaches to gene transfer will first be discussed, after which some representative examples of transgenic experiments for studying gene regulation and creating models of disease will be presented. Throughout this article, issues of reproductive management, aspects of colony maintenance that require special emphasis, and IACUC review of transgenic protocols will serve as themes. However, in the last segment of the manuscript, direct focus on specific aspects of laboratory animal management will be provided.

TRANSGENIC TECHNIQUES

Three approaches to genetic transformation of the mammalian germ line are currently in use. These are DNA microinjection, retrovirus-mediated gene transfer, and embryonic stem (ES) cell mediated gene transfer. Important features of all gene transfer approaches are their efficiency, the characteristics of foreign DNA processing during the gene transfer process, the integration mechanism itself, and characteristics of donor gene expression.

DNA Microinjection

DNA microinjection was the first transgenic technique demonstrated to be effective in mammals (Gordon and others 1980; Gordon and Ruddle 1981, 1983). With this approach, one-celled fertilized eggs are microinjected with purified DNA, and surviving embryos are then reimplanted into the uteri of females that are hormonally and physiologically competent to carry a pregnancy (Figure 1). While these females are usually not genetically related to the embryos received, there is no biological barrier to reimplantation of injected zygotes into female mice from which the progenitor oocytes are retrieved, as might well be the case in the event such procedures were performed in primates. The major advantages of microinjection are its technical simplicity, its capability for allowing efficient foreign gene expression, and its applicability to a wide variety of species. It has been possible to produce transgenic animals of virtually any species in which pronuclei can be visualized (Hammer and others 1985).

The efficiency of transgenic mouse production by pro-nuclear microinjection varies with different DNA constructs, with an average success rate of about 30% of live born animals. We have observed frequencies of integration as high as 90% for the plasmid pBR322 digested with PstI prior to microinjection. In addition to demonstrating the variability of DNA integration rate, this finding also underscores the point that the coding information contained within the donor DNA does not obviously affect the frequency of transgenic animal production: pBR322 is a bacterial sequence with no significant homology to mammalian DNA or coding information recognizable by a mammalian cell. The frequency of birth of transgenic pups can be reduced if pathogenic genes such as expressible oncogenes or toxin genes are inserted. Under these circumstances expression during embryonic or fetal life could kill embryos. Another factor relating to efficiency of the process, which influences the number of female animals used as donors of embryos for microinjection, is that not all embryos survive microinjection, with 90% survival obtained under optimal conditions. In our laboratory, we generally superovulate 8-10 females to obtain embryos for one experiment, and we usually microinject 3 times per week, 48 weeks of the year. IACUC protocols that accompany our grants accordingly request the use of 1,440 donor females each year.

The process of foreign gene integration is poorly understood, but usually leads to insertion within the genome of multiple copies of the donor DNA fragment arranged in head-to-tail arrays (Costantini and Lacy 1981). Not only is it currently impossible to control the number of gene copies that ultimately integrate, but the integration event can be associated with significant rearrangement of host DNA sequences. For example, we have observed a translocation in a line of transgenics carrying a human interferon gene, with the foreign gene mapping to one of the translocation chromosomes (Gordon and others 1989). Another important aspect of integration after microinjection is that it is apparently random with regard to insertion site. This fact creates the risk that host genes will be interrupted by insertion of new DNA, and as a consequence, that host gene function will be negated. Thus new, "insertional mutations", some with significant pathological consequences, can be introduced during gene transfer (Woychik and others 1985; Krulewski and others 1989). The "risk" of insertional mutagenesis is more appropriately regarded as an opportunity, since genes mutated by insertion of a defined DNA element are subject to molecular cloning. As a consequence genes with important roles in development can be isolated, and the relationship between their expression and normal mammalian development elucidated. For example, a mutation that causes situs inversus in mice has tentatively been isolated (Singh and others 1991). It should be appreciated that insertional mutations are irreplaceable, since they result from a chance integration event. As such, these animals are extremely valuable, and investigators can be expected to breed such lines aggressively in order to ensure their preservation.

Expression of foreign genes is also an important aspect of the gene transfer strategy. Microinjected genes can express efficiently, but the level of expression can be influenced by several factors. This point is illustrated by the history of globin gene expression in transgenic mice:

The globin gene cluster contains a group of genes that are sequentially activated during development, with b-globin expressed in the adult bone marrow. When a relatively large (about 15 kilobase) DNA fragment containing the rabbit b-globin gene cloned into bacteriophage l was first inserted into transgenic mice, no expression in bone marrow was detectable (Lacy and others 1983). However when smaller fragments encompassing the human gene were excised from a plasmid and microinjected, tissue specific, developmentally regulated expression of the globin gene was observed (Chada and others 1985; Townes and others 1985)--that is, expression occurred in fetal liver and adult bone marrow. However, the level of expression was far lower than that of endogenous mouse b-globin, and varied substantially between different lines of animals that harbored the transgenes at different chromosomal locations. These experiments showed that prokaryotic (bacteriophage or plasmid) sequences can inhibit transgene expression, and further indicated that small fragments of the b-globin gene, while expressed properly with regard to spatial and temporal regulation, were not expressed at levels proportional to the number of gene copies present. Later it was shown that regulatory elements as far as 50 kilobases (kb) from the globin coding sequence, identified by their presence within DNAseI hypersensitive chromatin, were important to regulation of the gene, and when these enhancers were included on new transgene constructs, copy number dependent, integration site independent expression was seen (Grosveld and others 1987). Thus, although b-globin genes are active in only a single adult tissue, they are regulated by multiple elements. These findings have important bearing on the design of transgenic experiments directed toward characterizing gene regulatory elements. It is not unusual for such protocols to involve the production of many transgenic lines with a variety of different gene constructs.

Retrovirus-mediated Gene Transfer

Little time will be devoted here to retrovirus-mediated gene transfer, which is based upon the finding of Jaenisch (1976) that retroviruses can infect early embryos and insert proviral DNA copies of their RNA genomes into the mouse chromosome (Figure 2). This system is technically simple, requiring only exposure of zona-free embryos to medium containing recombinant retroviruses. The system has the additional advantages that integration causes minimal disruption of host DNA and always involves integration of a single copy of the donor gene. Recombinant retroviruses are produced by packaging the replication-defective, recombinant genomes in packaging cell lines that produce retroviral proteins but are unable to package wild type, potentially oncogenic viral genes (Mann and others 1983). This gene transfer methodology is not commonly used because gene expression in such viral constructs is often inefficient, and the size of the viral genome limits to about 9 kb the amount of heterologous genetic material that can be inserted. In contrast to this size limitation, it has recently been shown that microinjection of embryos or transfection of embryonic stem cells (see below) can introduce DNA fragments of several hundred thousand bases when cloned in yeast artificial chromosomes (Schedl and others 1992, 1993; Choi and others 1993; Lamb and others 1993).

Concern has periodically been raised that retroviral transgenes could recombine with host DNA sequences so as to reconstruct an infectious virion. Original viral packaging extracts indeed exhibited this problem on occasion (Mann and others 1983). In these cases reconstruction of infectious virus took place during the packaging of the recombinant virus within the cytoplasm of the packaging cells, which harbored retroviruses that were lacking only the packaging signal, or Y sequence. In order to solve this problem, packaging cell lines were developed in which genes encoding various retroviral proteins were cloned onto separate plasmids and transferred into separate chromosomal sites (Markowitz and others 1988). This procedure effectively eliminates generation of infectious virions. It should also be recognized that infectious retroviruses within laboratory animals pose no apparent danger to animal care personnel, as the viruses are unable to infect human cells.

Embryonic Stem Cell Mediated Gene Transfer

Embryonic stem (ES) cell mediated gene transfer is a complex but powerful approach to germline gene insertion. Thus far established only from mice, embryonic stem (ES) cells are derived from what is presumed to be the inner cell mass component of the blastocyst. The remarkable feature of these cells is that after extended periods in culture they can be inserted into the cavities of normal mouse blastocysts and induced to resume a normal program of development, whereupon they differentiate into all cell types of an adult mouse including germ cells (Evans and Kaufman 1981; Martin 1981). The mouse initially produced from such cell injections is usually a genetic chimera composed of ES derivatives as well as cellular descendants of the blastocysts into which the ES cells were inserted. An important key to success of ES cell technology is the production of germ cells from the cellular component of the chimera that is derived from the cultured, genetically engineered ES cells (Figure 3).

The important attribute of these cells as vectors for genetic modification of the organism is based upon the discovery of Smithies and others (1985) that in tissue culture, donor gene constructs can replace corresponding segments of host DNA by homologous recombination. Mansour and others (1988) have subsequently devised efficient methods of selecting for this relatively rare form of gene insertion. Because gene substitutions can be made in this way, the opportunity to modify a host locus without affecting important flanking sequences exists, and as such, this method closely approximates the "perfect" gene transfer procedure. That is, a single copy of the new gene is inserted and can be regulated by all relevant enhancing elements at the appropriate genetic locus. The most common application of ES technology has thus far been to "knock out" genes by homologous recombination. With this procedure, DNA constructs that have significant sequence homology with a host gene are modified such that the coding element no longer functions. A typical modification is substitution of a bacterial drug resistance gene (for example neomycin resistance) for a portion of the coding region of the mammalian gene to be deactivated. When the donor DNA is applied to ES cells, insertion of the material into the genome can be identified by selection using neomycin, followed by detailed molecular analysis of the surviving clones. The neomycin resistance gene on the donor fragment allows successfully transfected cells to survive neomycin exposure. The efficiency of this process can be improved by additionally including a negative selectable marker such as the herpes thymidine kinase gene, expression of which kills dividing cells in the presence of drugs such as ganciclovir. These negative selectable markers are lost during homologous recombination events, which simultaneously lead to insertion of the positive selectable marker for neomycin resistance. Consequently, ES cells that are resistant to both neomycin and ganciclovir are more likely to have sustained a homologous recombination event than randomly selected neomycin resistant clones (Mansour and others 1988).

Using this and related approaches it has been possible to create mice that lack specific genes. The deletion of a specific genetic function in all cells throughout development can yield fascinating insights into the genetic control of important developmental processes. For example, when the c-myb protooncogene is deleted, the developing fetus is unable to transfer the site of erythropoiesis from the embryonic yolk sac to the fetal liver, as normal animals do, and the fetuses die of anemia (Mucenski and others 1991).

Perhaps the most interesting feature of knockout mice is that mutations that correspond to those found in human disease often do not elicit the same pathology in mice as in humans. In humans, HPRT deficiency causes Lesch-Nyhan syndrome, a disorder characterized by mental retardation; chorioathetotic motor dysfunction; and bizarre, self-mutilating behavior. However, when this X-linked gene is deleted in mice, no abnormalities are observed (Hooper and others 1987; Kuehn and others 1987). Another interesting example is that of the retinoblastoma (Rb) gene. Mutations at this locus are associated with dominantly inherited retinoblastomas. Although individuals with retinoblastoma are typically heterozygous carriers of a mutation at Rb, the tumors themselves can be shown to have lost the remaining wild-type copy of the gene. Mice carrying a targeted mutation of the Rb locus have similar molecular changes in the tumors that arise, but the tumors occur in the pituitary gland rather than the retina (Lee and others 1992; Clarke and others 1992; Jacks and others 1992).

Discrepancies in patterns of pathology between humans with mutations and corresponding knockout mice have led some to suggest that these animals are not so valuable for disease modeling. This view is simplistic and short-sighted. While interspecies differences can alter specific patterns of development and such disturbances in those patterns can be different, the effects observed in mice are no less instructive with regard to the fundamental developmental mechanisms controlled by specific genes. A "library" of knocked out genetic loci, stored as frozen mouse embryos, would be an invaluable resource for future investigations of development and disease, as well as the function of complex gene regulatory networks.

ES cell mediated gene transfer is currently not applicable to species other than mice because it has not yet been possible to develop such cell lines in other species. Reports of the creation of ES lines from other species are now emerging, but it must be kept in mind when reading such literature that documentation of the existence of a line is critically dependent upon the demonstration that the cells will colonize the germ line. Even when this is demonstrated, ES cell technology may be difficult to apply to animals that produce few progeny and that have long gestation times. In mice, it is possible to quickly determine that an ES line cannot contribute to the germ line by breeding an animal and examining 10-50 of its offspring. However, in cattle, this critical genetic test would be quite difficult to perform. Chimera construction is not as practical an approach as nuclear transplantation, but extensive experimentation could be required to determine that nuclei from a specific ES line are not developmentally totipotent.

APPLICATIONS OF TRANSGENIC TECHNOLOGY

Transgenic animals, primarily mice, used thus far almost exclusively for basic research, have revolutionized our knowledge of mammalian developmental genetics. However, this technology also has significant potential for genetic engineering of livestock. The many uses of transgenic mice for basic studies have been reviewed previously (Gordon 1989). In the present paper, the use of gene insertion to elucidate mechanisms of gene regulation will be briefly reviewed, after which an example from our laboratory of a transgenic model of the human neurodegenerative disease amyotrophic lateral sclerosis will be described. Models of hepatitis B infection and ankylosing spondylitis will also be presented. These examples are discussed not only because they illustrate the power of this experimental paradigm, but because they highlight some of the ways in which transgenic experimentation can have an impact on laboratory animal management.

Transgenic Mice and Gene Regulation

Perhaps the most important contribution of transgenic technology to biomedical science has been its use for determining the mechanisms by which genes become differentially regulated during mammalian development. Since all cells of the developing mammal are clonal descendants of the fertilized egg, and since all contain the same genetic information, it is critical to understand how, during the process of differentiation, tissue-specific genes are activated while most others are permanently silenced. The ability to introduce new genes into random sites in the genome provided an important tool for solving this puzzle.

One of the early publications involving transgenic mice provided an important piece to that puzzle: Brinster and others (1981) linked the gene encoding Herpes virus thymidine kinase (TK) to several hundred bases of DNA that immediately flanked the mouse metallothionein-1 (MT-l) gene, and that contained the gene promoter. The MT-I gene, though ubiquitously expressed, is most active in liver, and is inducible by heavy metal exposure. When the MT- 1/TK construct was introduced into mice, transgenic animals expressed TK in a pattern typical of MT- 1, and the gene could be activated further by exposing the animals to cadmium. This study demonstrated that cis-acting elements in or around genes carry information that dictate tissue-specificity of expression. As noted earlier in the discussion of globin gene expression, these cis-acting elements are subject to effects of chromosomal position if they are incomplete. While some genes carry multiple elements (such as globins), others appear more simply regulated (MTs). Genes expressed in multiple different tissues contain many different elements, with some devoted to expression in each tissue (Gordon and others 1987). When genes are transferred across species barriers, evolutionary changes in organ-specific gene regulation can lead to aberrant patterns of transgene expression. For example, the human fetal globin gene, normally active in the human fetal liver, is expressed in the embryonic yolk sac in transgenic mice. This is because the mouse uses its own fetal gene for production of globin in the embryonic hematopoietic compartment (Chada and others 1986). Details of these regulatory phenomena are reviewed extensively elsewhere (Gordon 1989).

A Transgenic Mouse Model of Amyotrophic Lateral Sclerosis

One of the most important uses of transgenic animals is for modeling human diseases or disease processes. Among the major successes in this area has been the production of a transgenic mouse model of amyotrophic lateral sclerosis (ALS), also called Lou Gehrigs disease. ALS is an age-related neurodegenerative disorder that primarily involves motor neurons. Although the majority of ALS cases are sporadic, a subset of affected individuals inherit the disease. Rosen and others (1993) observed that a subset of individuals with familial, autosomal dominantly inherited ALS (FALS) harbor mutations of the Cu/Zn superoxide dismutase (SOD-l) gene. These findings indicate a causative relationship between altered SOD-1 activity and motoneuron degeneration. Because humans with ALS are generally not identified until muscle weakness sets in, early medical intervention and prevention is difficult. For these reasons it is highly desirable to develop an animal model of FALS. To this end, we introduced a mutation into the 4th exon of a 15 kb mouse genomic clone that we had previously isolated (Benedetto and others 1991) such that a glycine residue (GGC) at position 86 of the protein was changed to an arginine residue (CGC). This same mutation is found in some families with FALS (Rosen and others 1993). The sequence change not only creates a mouse counterpart of a pathogenic human gene sequence, it introduces a recognition sequence for FspI restriction endonuclease. The new restriction enzyme site allowed us to distinguish the transgene DNA and RNA from those of the endogenous mouse genes. This construct was then microinjected to produce transgenic mice (Ripps and others 1995).

In 2 mouse lines that produced high levels of transgene mRNA in the central nervous system (CNS), motor paralysis developed. This phenotype was associated histopathotogically with degenerative changes of motoneurons within the spinal cord, brain stem, and neocortex (Ripps and others 1995). These animals constitute a potentially valuable animal model of ALS. In humans with FALS, who harbor mutations at the SOD-1 locus, many genetic and environmental variables complicate analysis of the relationship between the mutations found and the appearance of this disease. The transgenic mice produced are of the highly inbred FVB/N strain and are housed under uniform conditions. As a consequence, appearance and progression of symptoms is very consistent between animals, such that all mice die between the ages of 94 and 117 days. This consistency allows for sensitive testing of potential therapeutic agents or environmental factors that might alleviate or exacerbate the disease, because a change in lifespan of only a few days would be statistically significant. Moreover, because the genetic basis for the disease can be known immediately after birth, early manifestations of the disorder can be studied in detail, and therapeutic interventions can be attempted at all stages of disease progression. Similar findings have been reported by a group that introduced a mutated human gene into the mouse germ line (Gurney and others 1994).

Important features of these abnormal animals from the point of view of laboratory animal management is early onset and rapid progression of the paralysis, as well as reduced breeding efficiency. Females breed very poorly, and males die at 94-117 days of age. These characteristics require that the transgene be transmitted through males in the few weeks between the onset of reproductive maturity and death from motor paralysis, in order to assure survival of the line, we place a male with several females. When a female delivers pups, she remains in the cage for at least one additional day to assure that she mates again. Only then is the female and litter separated from the other adults in the group. These kinds of breeding practices are unusual, but often necessary, and tolerance of them by veterinary personnel and others should be encouraged.

Transgenic Models of Hepatitis B Infection

Chronic hepatitis B (HBV) infection is one of the most common gastrointestinal diseases in the world, and can have serious sequelae: Chronic active hepatitis, which is associated with ongoing liver cell injury, dramatically increases the risk of developing hepatocellular carcinoma (HCC). However, the mechanism by which chronic hepatitis B infection predisposes to development of HCC remains unclear. The chronic inflammatory process, accompanied by liver cell death and regeneration, may ultimately lead to transforming mutations in hepatocytes. Alternatively integration of viral DNA, a common event in HCC cells, may disturb host cell gene regulation and lead to malignant degeneration.

Because of the narrow host range of this virus, there are very few useful animal models of HBV infection. However, transgenic technology affords the opportunity to produce mouse models of the condition. Moreover, because transgenic technology inevitably involves integration of the donor DNA, this particular characteristic of HBV DNA in HCC can readily be achieved in transgenic mice.

Initial efforts to produce HBV transgenic mice focused on the surface antigen gene (HBsAg), the prevalent circulating antigen in chronic persistent hepatitis. When this gene was linked to the MT-I promoter (Chisari and others 1985), introduced under control of its own promoter (Babinet and others 1985), or, if the entire HBV genome was introduced (Farza and others 1988; Araki and others 1989), production of HBsAg and antigenemia could be documented. Mice expressing surface antigen were found to be more sensitive to chemical carcinogenesis in the liver (Dragani and others 1989; Sell and others 1991), findings that suggest that environmental factors may play an important role in the transition from chronic infection only to HCC.

When Chisari and others (1989) linked the large envelop polypeptide region, which contains the HBsAg gene, to the strong albumin promoter/enhancer region, transgenic mice expressed large quantities of HBsAg and eventually developed HCC (Chisari and others 1989; Dunsford and others 1990). These findings support the hypothesis that chronic liver cell injury plays an important role in HBV-induced HCC (Dunsford and others 1990).

Another HBV gene implicated in HCC is HBx gene, which encodes a transcriptional trans activator. Kim and others (1991) introduced this gene under the control of its own promoter into mice, and observed high expression of HBx in liver, kidney, and testis. By 4 months of age focal areas of hepatocyte abnormalities were present, and definite tumors diagnosed as clear cell carcinomas were manifest by 8 to 10 months. Carcinomas, some with metastases, killed the animals by 11 to 15 months of age.

In summary, the species barrier to HBV infection can be overcome in mice by direct microinjection of HBV genes. The genes integrate, and when HBsAg is expressed at high levels, liver cell injury and HCC develop. The HBx gene alone, when expressed at high levels, also can cause liver cancer. It is unclear whether any of these HBV transgenic mice develop liver cancer by the same mechanism as humans.

To what extent do transgenic mice carrying the genes of a dangerous human pathogen such as HBV pose a threat of zoonotic transmission of disease? Of all HBV transgenic mice, only those that carry the entire HBV genome have the potential to produce an infectious HBV particle. Although HBV cannot infect mouse cells, it is possible that an animal with the entire genome in every cell could produce mature virions and release them into its serum. In fact, some transgenic mice that carry the entire HBV genome carry core antigens in blood (Araki and others 1989). While viral titers are extremely low, it must be acknowledged that in cases such as this, the risk of zoonoses may be significant.

To date, transmission of disease specific to expression of a transgene from a transgenic animal to man has not been reported. However, it is important to study new proposals involving insertion of human pathogens to determine if risk exists.

Transgenic Rats with Ankylosing Spondylitis

In humans, the HLA-B27 haplotype is associated with inflammatory disease involving the skin, gastrointestinal tract, genitourinary tract, eye, and heart. The disorder with the strongest link to HLA-B27 is ankylosing spondylitis, an inflammatory disorder that most often causes inflammation of the spine or sacroiliac joints, or both. A possible causative role for the HLA-B27 haplotype and development of the disease was not established until Hammer and others (1990) produced transgenic rats carrying human HLA-B27 determinants. In this experiment the HLAB27 genes were co-injected with the gene encoding human b₂-microglobulin, which is required for MHC antigen presentation. Transgenic rats expressing high levels of HLA-B27 on the cell surface in association with b₂-microglobulin developed a variety of inflammatory disorders that strongly resembled human HLA-B27-related disease, including gastrointestinal, skin, genitourinary joint, and cardiac involvement. These animals thus establish a model system for studying many of the perplexing aspects of these disorders, including the fact that many of the pathological sequelae are precipitated by an episode of infection.

This model is mentioned here not only because it constitutes a major breakthrough in the understanding of these inflammatory diseases, but because it illustrates that different species can respond differently to transgene expression. In prior experiments where HLA-B27 genes were expressed in transgenic mice, no pathology resulted (Hammer and others 1990), while the transgenic rats exhibited profound phenotypic changes. This observation is relevant to the evaluation by animal care committees of proposed transgenic experiments. Selection of species for gene insertion can be critical to an experiment and thus, the proposal to insert genes into species other than mice should not be regarded as excessive or unnecessary without prior thorough analysis.

TRANSGENIC EXPERIMENTS AND LAB ANIMAL MANAGEMENT

Although not directly related to micromanipulation technique, transgenic colony management, especially at a time when animal rights activism is increasing animal care costs, is a very important skill. In this section, aspects of transgenic technology that are relevant to evaluation of proposals and to management of laboratory animals are discussed. Recently the Institute of Laboratory Animal Resources (ILAR) has issued a document on rodent management which includes a detailed discussion of transgenic animals (NRC 1993).

Establishing and Maintaining Lines Requires Aggressive Breeding

When transgenic founders are first identified, it is important to breed them aggressively until the line is established, because the presence and expression of a transgene can compromise fertility and because the founder may be a mosaic, with the transgene present in only a subset of germ cells. For these reasons, lab animal care personnel should be prepared to provide the necessary equipment and space.

Embryo Freezing

When freezing embryos for transgenic strain preservation, it is important to realize that large numbers of embryos are required to assure rapid reconstitution of the strain. These numbers may be augmented further if the transgenic line cannot be bred to homozygosity. The inability to produce fertile homozygotes of both sexes is not uncommon, and it is due to insertional mutagenesis, chromosomal rearrangement associated with transgene integration, or to physiological changes that render the animal infertile (for example, growth hormone expression from hepatic transgenes in females). Embryo freezing also requires a backup system to secure against freezer breakdown, and the availability of skilled labor. Establishment of a freezing program, therefore, requires significant cost, an adequate organizational infrastructure, and time for strain reconstitution. Commercial freezing is available but can be costly. For these reasons, it is often desirable to maintain small samples of breeding nuclei for various strains "on the shelf." When homozygotes can be produced this strategy is simple and cost-effective, as DNA analysis in the line becomes unnecessary. Tests for transgene homozygosity must be rigorous, however, in order to assure that the transgene is not lost when DNA screening is discontinued. Testing should be done both by Southern blotting using a reference probe to track the amount of DNA loaded, and by genetic testing. Once putative homozygotes are produced and crossed with each other, they should be watched carefully for reproductive competence, as homozygotes of either sex may be infertile.

Cost Containment

As animal purchase and maintenance costs steadily increase while grant funds decrease, cost containment of transgenic colonies becomes very important. Perhaps the most important cost-saving measure is effective design of transgenes. When the sequence to be microinjected is poorly designed, mice are produced that give less than the expected result. This situation frequently leads to new rounds of microinjection with new constructs, but few investigators discard the previous mice. Therefore, because the effort to produce transgenic mice is substantial and because animal care costs are high, it is advisable to plan experiments carefully. When several lines are produced with a single construct, it may be advisable to discontinue some of the lines. This is psychologically difficult because of the potential loss of data but it may prove to be unavoidable. Obviously it is important to characterize the phenotypes of the various lines before making such a decision.

A third measure toward cost containment involves reproductive management. Since male mice can be maintained for a significant period of time without losing fertility, it is possible to maintain some lines by simply keeping 1-2 males on the shelf for up to 6 months without breeding them, provided that the males are known to transmit the transgene effectively. The inclination to expand every line to large numbers should be avoided; instead, a rational plan should be developed concerning the number of animals, both positive and negative, that will actually be needed for each experiment. Although establishment of homozygous transgenic mice can be difficult and is labor intensive, it is, in the long-term, a highly effective cost-saving measure. When every animal in a line is transgenic, DNA screening can be halted, and very few animals are needed to ensure preservation of the line. Animal care personnel should be sensitive to the burdensome cost of transgenic experiments, and should assist investigators in development of management strategies that minimize costs. It must also be recognized that if current trends continue, the high cost of per diem animal maintenance will necessitate institutional support for transgenic colonies, since such funding is not completely covered by grants from the National Institutes of Health.

Ethics of Transgenic Technology

It is not uncommon for animal care committees to confront questions relating to the ethics of transgenic technology. The transgenic process has been criticized from several perspectives. First it has been suggested that escape of transgenic animals could pose a unique threat to wild animal populations by introducing foreign genes into those populations through breeding. Second, animals expressing pathogenic transgenes may suffer from unique diseases. Third, concern has been raised that transgenic animals increase animal use. Finally, the very notion of performing genetic engineering experiments has been criticized.

Although it is not possible to be certain that escape of transgenic animals will not genetically contaminate wild animal populations, these concerns are probably not serious. Not only are escapes from well-managed facilities unlikely, but laboratory animal strains are far less vigorous and capable of surviving in the wild than are their feral counterparts. Moreover, it is highly unusual for foreign genes to confer a selective advantage upon a transgenic animal. Under most circumstances new genes are either neutral with respect to animal health or are deleterious. Thus, for the most part, transgenic animals are especially dependent upon the laboratory environment for survival and reproduction.

While expression of genes that cause tumors or neurodegenerative diseases undoubtedly causes animal suffering, it should be recognized that similar pathologies have historically been caused by selective breeding of laboratory animals. The vast majority of commercially available mutations which cause such disease have arisen spontaneously in breeding colonies. Just as the value of these mutations to science has never been questioned, the value of disease models related to transgene expression cannot be questioned. Indeed, transgenic models are often especially valuable because their genetic basis is understood.

The concern that animal use will increase as transgenic technology is promulgated may be simplistic. Of course, when gene constructs are tested for tissue distribution of expression, or when disease models are sought, transgenic approaches are likely to be pursued, and more animals will be used. However, because transgenic technology is so efficient and powerful, results may be obtained more quickly, and ultimately, after the use of fewer animals. Moreover, the establishment of transgenic mouse substitutes for models previously involving larger animals can reduce the use of animals that pose greater ethical challenges. A good example is the development of transgenic mice for poliovirus vaccine testing, which previously has required monkeys (Ren and others 1990; Koike and others 1991). It is also my view that care should be taken to resist the current psychological drift from minimization of animal use for valid experiments, to elimination of valid experiments for the sole purpose of minimizing animal use. This subtle difference, if not recognized, could severely hamper biomedical research.

Whether genetic engineering itself is ethical is of course a philosophical question. However, it should be appreciated that genetic engineering is an old practice, which has recently added transgenic technology to its traditional armamentarium of procedures. Selective breeding, the most widely used form of genetic engineering techniques, is in many respects more powerful than transgenic technology. If a phenotype such as rapid growth is sought by selective breeding, relevant alleles of a large number of genes that control this process can be assembled in a single animal strain. Candidate genes could regulate growth hormone production, end-organ responsiveness to growth hormone, skeletal and muscle structure, and other factors. As a consequence of this, pedigreed animals bear little resemblance to their ancestors or feral counterparts. Transgenic technology is relatively haphazard in the sense that a single gene is selected for modification of a possibly complex phenotypic trait. It is probably for this reason that efforts to modify livestock through transgenic technology have not yet been highly rewarding. Transgenic technology as a genetic engineering tool should therefore be regarded in the proper perspective. It is rapid, in the sense that genes from widely disparate species can be brought together in a single generation, but it is limited by the technical reality that relatively few traits can be modified in any single experiment.

SUMMARY AND CONCLUSIONS

Transgenic technology is one of the most profound scientific advances in the recent history of developmental genetics. It provides the investigator with enormous power to develop models of disease, models for disease treatment and prevention, and genetic modification of livestock. This advance has already benefited human health by providing deep insights into mechanisms of gene regulation, with aberrancies of the latter being responsible for a wide variety of human diseases. This technology will certainly become more important in the future, as variations of the techniques, and applications of them to other species, are developed and promulgated. It is important for professionals in the laboratory animal sciences to appreciate the importance of this work and the special problems it poses both to investigators and animal facilities.

ACKNOWLEDGMENTS

Work toward preparation of this manuscript was supported by NIH grants HD20484 and AGO6647.

REFERENCES

Araki K, Miyazaki J, Hino O, Tomita N, Chisaka O, Matsubara K, Yamamura K. 1989. Expression and replication of hepatitis B virus genome in transgenic mice. Proc Nail Acad Sci USA 86:207-211.

Babinet C, Farza H, Morello D, Hadchouel M, Pourcel C. 1985. Specific expression of hepatitis B surface antigen (HBsAg) in transgenic mice. Science 230:1160-1163.

Benedetto MT, Anzai Y, Gordon JW. 1991. Isolation and analysis of the mouse genomic sequence encoding Cu/Zn superoxide dismutase. Gene 99:191-95.

Brinster RL, Chen HY, Trumbauer ME, Senear AW, Warren R, Palmiter RD. 1981. Somatic expression of herpes thymidine kinase in mice following injection of a fusion gene into eggs. Cell 27:223-231.

Chada K, Magram J, Raphael K, Radice G, Lacy E, Costantini F. 1985. Specific expression of a foreign (-globin gene in erythroid cells of transgenic mice. Nature 314:377-80.

Chada K, Magram J, Costantini F. 1986. An embryonic pattern of expression of a human fetal globin gene in transgenic mice. Nature 319:685-689.

Chisari FV, Pinkerr CA, Milich DR, Fillippi P, McLachlan A, Palmiter RD, Brinster RL. 1985. A transgenic mouse model of the chronic hepatitis B surface antigen carrier state. Science 230:1157-1160.

Chisari FV, Klopchin K, Moriyarna T, Pasquinelli C, Dunsford HA, Sell S, Pinken CA, Brinster RL, Palmiter RD. 1989. Molecular pathogenesis of hepatocellular carcinoma in hepatitis B virus transgenic mice. Cell 59:1145-1156.

Choi TK, Hollenbach PW, Pearson BE, Ueda RM, Weddell GN, Kurahara CG, Woodhouse CS, Kay RM, Loring J. 1993. Transgenic mice containing a human heavy chain immunoglobulin gene fragment clones in a yeast artificial chromosome. Nature Genet 1:17-123.

Clarke AR, Maandag ER, van Roon M., van der Lugt NMT, van der Valk M., Hooper M.L., Bems A., Riele H. 1992. Requirement for a functional Rb-1 gene in murine development. Nature 359:328-330.

Costantini F, E Lacy. 1981. Introduction of a rabbit b-globin gene into the mouse line. Nature 294:92-4.

Dragani TA, Manenti G, Farza H, Della Porta G, Tiollais P, Pourcel C. 1989. Transgenic mice containing hepatitis B virus sequences are more susceptible to carcinogen-induced hepatocarcinogenesis. Carcinogenesis 11:953-956.

DunsiBrd HA, Sell S, Chisari FV. 1990. Hepatocarcinogenesis due to chronic liver cell il/jury in hepatitis B virus transgenic mice. Cancer Res 50:3400-3407.

Evans MJ, Kaufman MH. 1981. Establishment in culture of pluripotential cells from mouse embryos. Nature 292:154-56.

Farza H, Hadchouel M, Scotto J, Tiollais P, Babinet C, Pourcel C. 1988. Replication and gene expression of hepatitis B virus in a transgenic mouse that contains the complete viral genome. J Virol 62:4144-4152.

Gordon JW. 1988. Transgenic animals--the state of the art. ILAR News 30:8-18.

Gordon JW. 1989. Transgenic animals. Int Rev Cytol 115:171-230.

Gordon JW, Ruddle FH. 1981. Integration and stable germ line transmission of genes injected into mouse pronuclei. Science 214:1244-1246.

Gordon JW, Ruddle FH. 1983. Gene transfer into mouse embryos. Production of transgenic mice by pronuclear injection. Methods Enzymol. 101:411-433.

Gordon JW, Scangos GA, Plotkin DJ, Barbosa JA, Ruddle FH. 1980. Genetic transformation of mouse embryos by microinjection of purified DNA. Proc Natl Acad Sci USA 77:7380-7384.

Gordon JW, Chesa PG, Nishimura H., Rettig WJ, Maccari JE, Endo T, Seravalli E, Seki T, Silver J. 1987. Regulation of Thy-1 gene expression in transgenic mice. Cell 50:445-452.

Gordon JW, Pravtcheva D, Poorman PA, Moses MJ, Brock WA, Ruddle FH. 1989. Association of a foreign DNA sequence with male sterility and a translocation in a line of transgenic Mice. Somatic Cell Mol Genet. 15:569-578.

Grosveld F, Van Assendelft GB, Greaves GR, Kollias G. 1987. Position-independent, high-level expression of the human b-globin gene in trans-genic mice. Cell 51:975-985.

Gurney ME, Puk H, Chiu AU, Dal Camto MC, Polchow CY, Alexander DD, Caliendo J, Hantati A, Kwon YW, Deng H-X., Chen W, Zhai P, Sufit RL, Siddique T. 1994. Motor neuron degeneration in mice that express a human Cu/Zn superoxide dismutase mutation. Science 264: 1772-1775.

Hammer RE, VG Pursel, CE Rexroad Jr, Wall R J, Bolt DJ, Ebert KM, Palmiter RD, Brinster RL. 1985. Production of transgenic rabbits, sheep and pigs by microinjection. Nature 315:680-683.

Hammer RE, Maika SD, Richardson JA, Tang J-P, Taurog JD. 1990. Spontaneous inflammatory disease in transgenic rats expressing HLA-B27 and human p2M: An animal model of HLA-B27-associated human disorders. Cell 63:1099-1112.

Hooper M, Hardy K, Handyside A, Hunter S, Monk M. 1987. HPRT-deficient (Lesch-Nyhan) mouse embryos derived from germline colonization by cultured cells. Nature 326:292-295.

Jacks T, FazeIi A, Schmitt EM, Bronson RT, Goodell MA, Weinberg RA. 1992. Effects of an Rb mutation in the mouse. Nature 359:295-300.

Jaenisch R. 1976. Germ line integration and Mendelian transmission of the erogenous Moloney leukemia virus. Proc Natl Acad Sci USA 73:1260-1264.

Kim C-M, Koike K, Saito I, Miyamura T, Jay G. 199 I. HBx gene of hepatitis B virus induces liver cancer in transgenic mice. Nature 351:317-320.

Koike S, Taya C, Kurata T, Abe S, lse 1, Yonekawa H, Nomoto A. 1991. Transgenic mice susceptible to poliovirus. Proc Natl Acad Sci USA 88:951-955.

Krulewski TF, Neumann PE, Gordon JW. 1989. Insertional mutation in a transgenic mouse pedigree allelic with pod. Proc Natl Acad Sci USA 86:3709-3712.

Kuehn MR, Bradley A, Robertson EJ, Evans MJ. 1987. A potential animals model for Lesch-Nyhan syndrome through introduction of HPRT mutations into mice. Nature 326:295-298.

Lacy E, Roberts S, Evans EP, Burtenshaw MD, Costantini FD. 1983. A foreign (-globin gene in transgenic mice: integration at abnormal chromosomal position and expression in inappropriate tissues. Cell 34:343-358.

Lamb BT, Sisodia SS, Lawler AM, Slunt HH, Kitt CA, Kearns WG, Pearson PL, Price DL, Gearhart JD. 1993. Introduction and expression of the 400 kilobase precursor amyloidprotein gene in transgenic mice. Nature Genet 5:22-30.

Lee EY-H, Chang CY, Hu N, Wang Y-C J, Lai C-C, Herrup K, Lee W-H., Bradley A. 1992. Mice deficient for Rb are nonviable and show defects in neurogenesis and haematopoiesis. Nature 359:288-294.

Mann JR, Mulligan RC, Baltimore D. 1983. Construction of retrovirus packaging mutant and its use to produce helper-free defective retrovirus. Cell 33:153-159.

Mansour SL, Thomas KR, Capecchi MR. 1988. Disruption of the protooncogene int-2 in mouse embryo-derived stem cells: a general strategy for targeting mutations to non-selectable genes. Nature 336:348-352.

Markowitz D, Goff S., Bank A. 1988. A safe packaging line for gene transfer: Separating viral genes on two different plasmids. J Virol 62:1120-1124.

Martin GR. 1981. Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci USA 78:7634-7638.

Mucenski ML, McLain K, Kier AB, Swerdlow SH, Schreiner CM, Miller TA, Pietryga DW, Scott WJ Jr, Potter SS. 1991. A functional c-myb gene is required for normal murine fetal hepatic hematopoiesis. Cell 65:677-689.

[NRC] National Research Council. 1993. Definition, Nomenclature, and Conservation of Rat Strains. Washington DC: National Academy Press.

Palmiter RD, Brinster RL. 1986. Germ-line transformation of mice. Ann Rev Genet 20:465-499.

Ren R, Costantini F, Gorgacz EJ, Lee J J, Racaniello VR. 1990. Transgenic mice expressing a human poliovirus receptor: a new model for poliomyelitis. Cell 63:353-362.

Ripps ME, Huntley GW, Hof PR, Morrison JH, Gordon JW. 1995. Transgenic mice expressing an altered murine superoxide dismutase gene provide an animal model of amyotrophic lateral sclerosis. Proc Natl Acad Sci USA 92:689-693.

Rosen DR, Siddique T, Patterson D, Figlewicz DA, Sapp P, Hentati A, Donaldson A, Goto J, O'Regan JP, Deng H-X. and others. 1993. Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 362:59-62.

Schedl A, Beermann F, Thies E, Montoliu L, Kelsey G, Schutz G. 1992. Transgenic mice generated by pronuclear injection of a yeast artificial chromosome. Nucleic Acids Res 20:3073-3077.

Schedl A, Montoliu L, Kelsey G, Schutz G. 1993. A yeast artificial chromosome covering the tyrosinase gene confers copy number-dependent expression in transgenic mice. Nature 362:258-261.

Sell S, Hunt JM, Dunsford HA, Chisari FV. 1991. Synergy between hepatitis B virus and chemical hepatocarcinogens in transgenic mice. Cancer Res 51:1278-1285.

Singh G, Supp DM, Schreiner C, Mcaneish J, Marker J-J, Copeland NG, Jenkins NA, Potter SS, Scott W. 1991. Legless insertional mutations: morphological, molecular and genetic characterization. Genes Develop 5:2245-2255.

Smithies O, Grett RG, Boggs SS, Koralewski MA, Kucherlapati RS. 1985. Insertion of DNA sequences into the human chromosomal b-globin locus by homologous recombination. Nature 317:230-34.

Townes TM, Lingrel JB, Chen HY, Brinster RL, Palmiter RD. 1985. Eryth-roid-specific expression of human b-globin genes in transgenic mice. EMBO J 4:1715-1723.

Woychik RP, Stewart TA, Davis LG, D'Eustachio PD, P Leder. 1985. An inherited limb deformity created by insertional mutagenesis in a transgenic mouse. Nature 318:36-40.

FIGURE 1 Production of transgenic mice by pro-nuclear microinjection. Female mice are superovulated and mated, and a pronucleus of the fertilized egg is microinjected with DNA (top right). Surviving zygotes are reimplanted (center), and newborn pups are tested by Southern blotting for incorporation of new genes (bottom). Coat color markers are use to identify microinjected embryos.

FIGURE 2 Retrovirus-mediated gene transfer. Zona-free mouse embryos are cultured in the presence of virus-producing cell lines. Viruses integrate into the embryo (middle) and can be found in all cells of resultant offspring (bottom).

FIGURE 3 Embryonic stem cell mediated transgenesis. Embryonic stem cells (here called embryonal stem cells) are cultured and subjected to DNA-mediated gene transfer (DMGT, top) or other manipulations. Cells with the desired genetic alteration are then inserted into blastocyst cavities, whereupon they resume normal development and produce a genetically mosaic mice, with some cells derived from the ES line, and the others from the injected blastocyst (center). Founders, usually males, are then bred to pass on the gene, provided the ES cells differentiate into sperm. The F1 hybrid then carries the new gene in all cells (bottom).