Early postimplantation embryo lethality due to DNA rearrangements in a transgenic mouse strain

Contemporary Transposon Tools: A Review and Guide through Mechanisms and Applications of Sleeping Beauty, piggyBac and Tol2 for Genome Engineering

Transposons are mobile genetic elements evolved to execute highly efficient integration of their genes into the genomes of their host cells. These natural DNA transfer vehicles have been harnessed as experimental tools for stably introducing a wide variety of foreign DNA sequences, including selectable marker genes, reporters, shRNA expression cassettes, mutagenic gene trap cassettes, and therapeutic gene constructs into the genomes of target cells in a regulated and highly efficient manner. Given that transposon components are typically supplied as naked nucleic acids (DNA and RNA) or recombinant protein, their use is simple, safe, and economically competitive. Thus, transposons enable several avenues for genome manipulations in vertebrates, including transgenesis for the generation of transgenic cells in tissue culture comprising the generation of pluripotent stem cells, the production of germline-transgenic animals for basic and applied research, forward genetic screens for functional gene annotation in model species and therapy of genetic disorders in humans. This review describes the molecular mechanisms involved in transposition reactions of the three most widely used transposon systems currently available (Sleeping Beauty, piggyBac, and Tol2), and discusses the various parameters and considerations pertinent to their experimental use, highlighting the state-of-the-art in transposon technology in diverse genetic applications.

Transgenesis and Animal Welfare

The two main techniques used in biomedical research for the production of transgenic animals have several implications for animal welfare in terms of the Three Rs of Russell & Burch. Some are intrinsic to the transgenic objectives, while others relate to the effects of mutations, transgene expression, associated methodologies, and husbandry or production systems. All of these actual and potential implications for animal welfare demand serious consideration within a broad ethical analysis of the technology. In the light, of the Three Rs, this may require a fundamental reappraisal of the processes by which such scientific procedures are approved.

Germline transgenesis in rodents by pronuclear microinjection of Sleeping Beauty transposons

We describe a protocol for high-efficiency germline transgenesis and sustained transgene expression in two important biomedical models, the mouse and the rat, by using the Sleeping Beauty transposon system. The procedure is based on co-injection of synthetic mRNA encoding the SB100X hyperactive transposase, together with circular plasmid DNA carrying a transgene construct flanked by binding sites for the transposase, into the pronuclei of fertilized oocytes. Upon translation of the transposase mRNA, enzyme-mediated excision of the transgene cassettes from the injected plasmids followed by permanent genomic insertion produces stable transgenic animals. Generation of a germline-transgenic founder animal by using this protocol takes ∼3 months. Transposon-mediated transgenesis compares favorably in terms of both efficiency and reliable transgene expression with classic pronuclear microinjection, and it offers comparable efficacies to lentiviral approaches without limitations on vector design, issues of transgene silencing, and the toxicity and biosafety concerns of working with viral vectors.

Strategies for selection marker-free swine transgenesis using the Sleeping Beauty transposon system

Swine transgenesis by pronuclear injection or cloning has traditionally relied on illegitimate recombination of DNA into the pig genome. This often results in animals containing concatemeric arrays of transgenes that complicate characterization and can impair long-term transgene stability and expression. This is inconsistent with regulatory guidance for transgenic livestock, which also discourages the use of selection markers, particularly antibiotic resistance genes. We demonstrate that the Sleeping Beauty (SB) transposon system effectively delivers monomeric, multi-copy transgenes to the pig embryo genome by pronuclear injection without markers, as well as to donor cells for founder generation by cloning. Here we show that our method of transposon-mediated transgenesis yielded 38 cloned founder pigs that altogether harbored 100 integrants for five distinct transposons encoding either human APOBEC3G or YFP-Cre. Two strategies were employed to facilitate elimination of antibiotic genes from transgenic pigs, one based on Cre-recombinase and the other by segregation of independently transposed transgenes upon breeding.

One-step knockin for inducible expression in mouse embryonic stem cells

Transgenesis enables the elucidation of gene function; however, constant transgene expression is not always desired. The tetracycline responsive system was devised to turn on and off transgene expression at will. It has two components: a doxycycline (dox)-controlled transactivator (TA) and an inducible expression cassette. Integration of these transgenes requires two transfection steps usually accomplished by sequential random integration. Unfortunately, random integration can be problematic due to chromatin position effects, integration of variable transgene units, and mutation at the integration site. Therefore, targeted transgenesis and knockin were developed to target the TA and the inducible expression cassette to a specific location, but these approaches can be costly in time, labor, and money. Here, we describe a one-step Cre-mediated knockin system in mouse embryonic stem cells that positions the TA and inducible expression cassette to a single location. Using this system, we show dox-dependent regulation of eGFP at the DNA topoisomerase 3β promoter. Because Cre-mediated recombination is used in lieu of gene targeting, this system is fast and efficient.

Homologous illegitimate random integration of foreign DNA into the X chromosome of a transgenic mouse line

Background

It is not clear how foreign DNA molecules insert into the host genome. Recently, we have produced transgenic mice to investigate the role of the fad2 gene in the conversion of oleic acid to linoleic acid. Here we describe an integration mechanism of fad2 transgene by homologous illegitimate random integration.

Results

We confirmed that one fad2 line had a sole integration site on the X chromosome according to the inheritance patterns. Mapping of insertion sequences with thermal asymmetric interlaced and conventional PCR revealed that the foreign DNA was inserted into the XC1 region of the X chromosome by a homologous illegitimate replacement of an entire 45,556-bp endogenous genomic region, including the ovarian granulosa cell tumourigenesis-4 allele. For 5' and 3' junction sequences, there were very short (3-7 bp) common sequences in the AT-rich domains, which may mediate the recognition of the homologous arms between the transgene and the host genome. In addition, analysis of gene transcription indicated that the transgene was expressed in all tested fad2 tissues and that its transcription level in homozygous female tissues was about twice as high as in the heterozygous female (p < 0.05).

Conclusions

Taken together, the results indicated that the foreign fad2 behaved like an X-linked gene and that foreign DNA molecules were inserted into the eukaryotic genome through a homologous illegitimate random integration.

Genetically engineering plants for crop improvement

Dramatic progress has been made in the development of gene transfer systems for higher plants. The ability to introduce foreign genes into plant cells and tissues and to regenerate viable, fertile plants has allowed for explosive expansion of our understanding of plant biology and has provided an unparalleled opportunity to modify and improve crop plants. Genetic engineering of plants offers significant potential for seed, agrichemical, food processing, specialty chemical, and pharmaceutical industries to develop new products and manufacturing processes. The extent to which genetically engineered plants will have an impact on key industries will be determined both by continued technical progress and by issues such as regulatory approval, proprietary protection, and public perception.

Efficient mammalian germline transgenesis by cis-enhanced Sleeping Beauty transposition

Heightened interest in relevant models for human disease increases the need for improved methods for germline transgenesis. We describe a significant improvement in the creation of transgenic laboratory mice and rats by chemical modification of Sleeping Beauty transposons. Germline transgenesis in mice and rats was significantly enhanced by in vitro cytosine-phosphodiester-guanine methylation of transposons prior to injection. Heritability of transgene alleles was also greater from founder mice generated with methylated versus non-methylated transposon. The artificial methylation was reprogrammed in the early embryo, leading to founders that express the transgenes. We also noted differences in transgene insertion number and structure (single-insert versus concatemer) based on the influence of methylation and plasmid conformation (linear versus supercoiled), with supercoiled substrate resulting in efficient transpositional transgenesis (TnT) with near elimination of concatemer insertion. Combined, these substrate modifications resulted in increases in both the frequency of transgenic founders and the number of transgenes per founder, significantly elevating the number of potential transgenic lines. Given its simplicity, versatility and high efficiency, TnT with enhanced Sleeping Beauty components represents a compelling non-viral approach to modifying the mammalian germline.

Animal pharming, two decades on

Since its inception 20 years ago, the animal pharming industry has promoted transgenic animals as a cost-effective method of biopharmaceutical production. However, it took until 2006 for the first therapeutic product to gain regulatory approval. This was an important milestone, but scepticism still abounds. Can pharming regain investor confidence, and will society accept transgenic livestock as a production method? There is some cause for optimism, biopharmaceuticals are a large, expanding market and animal pharming has already made considerable strides. A novel production platform has been established, groundbreaking technologies developed, a necessary regulatory framework put in place. Nevertheless, despite cost advantages, pharming has become a niche production method and its long term success may depend on products unique to transgenic animals.

Pigs taking wing with transposons and recombinases

Swine production has been an important part of our lives since the late Mesolithic or early Neolithic periods, and ranks number one in world meat production. Pig production also contributes to high-value-added medical markets in the form of pharmaceuticals, heart valves, and surgical materials. Genetic engineering, including the addition of exogenous genetic material or manipulation of the endogenous genome, holds great promise for changing pig phenotypes for agricultural and medical applications. Although the first transgenic pigs were described in 1985, poor survival of manipulated embryos; inefficiencies in the integration, transmission, and expression of transgenes; and expensive husbandry costs have impeded the widespread application of pig genetic engineering. Sequencing of the pig genome and advances in reproductive technologies have rejuvenated efforts to apply transgenesis to swine. Pigs provide a compelling new resource for the directed production of pharmaceutical proteins and the provision of cells, vascular grafts, and organs for xenotransplantation. Additionally, given remarkable similarities in the physiology and size of people and pigs, swine will increasingly provide large animal models of human disease where rodent models are insufficient. We review the challenges facing pig transgenesis and discuss the utility of transposases and recombinases for enhancing the success and sophistication of pig genetic engineering. 'The paradise of my fancy is one where pigs have wings.' (GK Chesterton).

Emergence of complex rearrangements at translocation breakpoints in a transgenic mouse; implications for mechanisms involved in the formation of chromosome rearrangements

Cryptic complex rearrangements as a result of a reciprocal chromosome translocation have been characterised in a transgenic mouse strain. Analysis of the breakpoint junctions in our previous studies showed that the ada transgene was integrated at the breakpoint forming a fusion gene with Golga3 (Mea2). In this study, further detailed analysis around the translocation junctions revealed that the surrounding regions were composed of 13 fragments of defined transgenic chromosome origins over approximately 1.9-Mb areas. Exactly the same cluster structure of these 13 breakpoint fragments already existed in the second generation of the transgenic mice. Our results show that this highly complex rearrangement has been conserved as the incipient form without any additional changes for 18 years up to the present generation, suggesting simultaneous occurrence of multiple events in the founder mouse.

Enzymatic engineering of the porcine genome with transposons and recombinases

Background

Swine is an important agricultural commodity and biomedical model. Manipulation of the pig genome provides opportunity to improve production efficiency, enhance disease resistance, and add value to swine products. Genetic engineering can also expand the utility of pigs for modeling human disease, developing clinical treatment methodologies, or donating tissues for xenotransplantation. Realizing the full potential of pig genetic engineering requires translation of the complete repertoire of genetic tools currently employed in smaller model organisms to practical use in pigs.

Results

Application of transposon and recombinase technologies for manipulation of the swine genome requires characterization of their activity in pig cells. We tested four transposon systems- Sleeping Beauty, Tol2, piggyBac, and Passport in cultured porcine cells. Transposons increased the efficiency of DNA integration up to 28-fold above background and provided for precise delivery of 1 to 15 transgenes per cell. Both Cre and Flp recombinase were functional in pig cells as measured by their ability to remove a positive-negative selection cassette from 16 independent clones and over 20 independent genomic locations. We also demonstrated a Cre-dependent genetic switch capable of eliminating an intervening positive-negative selection cassette and activating GFP expression from episomal and genome-resident transposons.

Conclusion

We have demonstrated for the first time that transposons and recombinases are capable of mobilizing DNA into and out of the porcine genome in a precise and efficient manner. This study provides the basis for developing transposon and recombinase based tools for genetic engineering of the swine genome.

Transgene constructs in coho salmon (Oncorhynchus kisutch) are repeated in a head-to-tail fashion and can be integrated adjacent to horizontally-transmitted parasite DNA

Currently, little information is available regarding the molecular organization of integrated transgenes in genetically-engineered fish. We performed a detailed structural analysis of an inserted transgene in one strain (M77) of transgenic coho salmon (Oncorhynchus kisutch) containing a salmon growth hormone gene construct (OnMTGH1). Microinjected DNA was found to have inserted into a single site in the coho salmon genome, and was organized with four complete internal copies and two partial terminal copies of the OnMTGH1 construct. All construct copies were organized in a direct-tandem (head-to-tail) repeat fashion in strain M77 and five additional strains (one also possessed a second recombinant junction fragment). For strain M77, the junctions between the transgene insert and the insertion point within the wild-type genome were cloned from strain-specific cosmid libraries and sequenced, revealing that the transgene insertion was accompanied by a deletion of 587 bp of wild-type DNA as well as a small insertion (19 bp) of unknown DNA upstream and a 14 bp direct- tandem duplication of sequence downstream. Upstream and downstream wild-type DNA sequence contained several repetitive sequence elements based on Southern blot analysis and homology to repetitive sequences in GenBank. In the downstream flank, a pseudogene sequence was also identified which has high homology to the CA membrane protein gene from Schistosoma japonicum, a parasite closely related to Sanguinicola sp. parasites which infect salmonids. Whether the presence of an inserted transgene and the presence of potentially horizontally-transmitted DNA are indicative of a genomic region with a predisposition for insertion of foreign DNA requires further study. The information derived from this transgene structure provides information useful for comparison to other transgenic organisms and for determination of the mechanism of transgene integration in lower vertebrates.

Characterization of transgene integration pattern in F4 hGH-transgenic common carp (Cyprinus carpio L.)

The integration pattern and adjacent host sequences of the inserted pMThGH-transgene in the F4 hGH-transgenic common carp were extensively studied. Here we show that each F4 transgenic fish contained about 200 copies of the pMThGH-transgene and the transgenes were integrated into the host genome generally with concatemers in a head-to-tail arrangement at 4-5 insertion sites. By using a method of plasmid rescue, four hundred copies of transgenes from two individuals of F4 transgenic fish, A and B, were recovered and clarified into 6 classes. All classes of recovered transgenes contained either complete or partial pMThGH sequences. The class I, which comprised 83% and 84.5% respectively of the recovered transgene copies from fish A and B, had maintained the original configuration, indicating that most transgenes were faithfully inherited during the four generations of reproduction. The other five classes were different from the original configuration in both molecular weight and restriction map, indicating that a few transgenes had undergone mutation, rearrangement or deletion during integration and germline transmission. In the five types of aberrant transgenes, three flanking sequences of the host genome were analyzed. These sequences were common carp b-actin gene, common carp DNA sequences homologous to mouse phosphoglycerate kinase-1 and human epidermal keratin 14, respectively.

Genetic Manipulation of Mammary Gland Development and Lactation

The mammalian genome is believed to contain some 30,000 to 40,000 different genes. Of these an estimated 42% have no known function. Genetically engineered mouse models (GEMM) have been a powerful tool available for determining gene function in vivo. In the mammary gland, a variety of genetic engineering approaches have been applied successfully to understanding the importance of specific gene products to mammary gland development and lactation. Our own laboratory has applied genetically engineered mice to facilitate understanding of the regulation of mammary gland development and lactation by insulin-like growth factors (IGF) and by the transcription factor, upstream stimulatory factor (USF-2). Our studies on transgenic mice that overexpress IGF-I have demonstrated the importance of IGF-dependent signaling pathways to maintenance of mammary epithelial cells during the declining phase of lactation. Our analysis of early developmental processes in mammary tissue from mice that carry a targeted mutation in the IGF-I receptor gene suggests that IGF-dependent stimulation of cell cycle progression is more important to early mammary gland development than potential antiapoptotic effects. Lastly, our studies on mice that carry a targeted mutation of the Usf2 gene have demonstrated that this gene is necessary for normal lactation and have highlighted the importance of this gene to the maintenance of protein synthesis. These studies, as well as studies of others, have highlighted both the strengths and limitations inherent in the use of GEMM. Limitations serve as the driving force behind development of new experimental strategies and genetic engineering schemes that will allow for a full understanding of gene function within the mammary gland.

Illegitimate DNA integration in mammalian cells

Foreign DNA integration is one of the most widely exploited cellular processes in molecular biology. Its technical use permits us to alter a cellular genome by incorporating a fragment of foreign DNA into the chromosomal DNA. This process employs the cell's own endogenous DNA modification and repair machinery. Two main classes of integration mechanisms exist: those that draw on sequence similarity between the foreign and genomic sequences to carry out homology-directed modifications, and the nonhomologous or 'illegitimate' insertion of foreign DNA into the genome. Gene therapy procedures can result in illegitimate integration of introduced sequences and thus pose a risk of unforeseeable genomic alterations. The choice of insertion site, the degree to which the foreign DNA and endogenous locus are modified before or during integration, and the resulting impact on structure, expression, and stability of the genome are all factors of illegitimate DNA integration that must be considered, in particular when designing genetic therapies.

High Toxicity, Low Receptor Density, and Low Integration Frequency Severely Impede the Use of Adenovirus Vectors for Production of Transgenic Mice

Although many methods are available for introducing genes into the mammalian germ line, none is ideal for genetic manipulation of livestock or primates. These organisms produce relatively few offspring in each reproductive cycle and they have long generation times. For these reasons, a recent report that adenovirus vectors can efficiently insert genes into the mouse germ line by embryo infection is of considerable interest. Adenovirus vectors have a high cloning capacity, can be produced in high titers, and can infect a wide variety of cell types. We have investigated in more detail the potential for such vectors to infect embryos and integrate their DNA into the genome. We exposed mouse embryos to adenovirus vectors that express bacterial beta-galactosidase (LacZ), and studied expression in the preimplantation period, toxicity of the vectors, and the frequency with which fetuses and pups integrate vector DNA. Our findings indicate that fully functional adenovirus receptor does not appear until the two-cell stage of development. Successful infection is associated with high toxicity, such that viral titers must be balanced to achieve high infection with tolerable levels of toxicity. Screening of 94 animals after embryo infection revealed a single positive polymerase chain reaction signal, which is indicative of the presence of the lacZ gene. This finding could not be confirmed by Southern blotting, which indicates that the founder animal was a genetic mosaic for the exogenous DNA. We conclude from these experiments that adenovirus gene transfer vectors are not readily usable for germ line gene insertion.

Cell-specific transgene expression from a widely transcribed promoter using Cre/lox in mice

Mice carrying two or more transgenes are used frequently to evaluate oncogene interactions during carcinogenesis. However, neoplastic transformation typically results in reduced expression both of differentiation-specific genes and of transgenes that use their promoters. In contrast, the more widely expressed metallothionein (MT) gene remains expressed at a high level in certain neoplasms, including those developing in pancreas. We have developed a system to maintain high-level, tissue-specific transgene expression during pancreatic carcinogenesis that uses Cre recombinase and a lox site-containing target transgene. Cre was expressed in pancreatic acinar cells under control of the elastase promoter (EL). Cre-mediated target transgene recombination placed a previously silent open-reading frame, encoding rat transforming growth factor alpha (TGFalpha), under control of the MT gene promoter. As long as DNA rearrangement does not occur in other cell types that express MT, TGFalpha expression will be restricted to acinar cells. Development of an effective target transgenic mouse required evaluation of multiple lineages to identify one with sufficient TGFalpha expression to induce pancreatic lesions after transgene rearrangement.

Cryopreservation of caprine beta-lactoglobulin transgenic mouse embryos

As an economical and safer alternative to the maintenance of transgenic animals as live stocks, transgenic embryo cryobanks can be generated and maintained indefinitely. Two-cell embryos obtained from four lines of caprine beta-lactoglobulin transgenic mice homozygous for the transgene were cryopreserved, and their response to cryopreservation-related stress was investigated. Significant differences between transgenic lines were found in the viability of frozen/thawed transgenic embryos and also in two-cell embryo production after superovulatory treatment of transgenic females. The results of this study suggest that cryopreservation protocols should be assessed on each transgenic line before the generation of transgenic embryo banks.

"Parahomologous" gene targeting in Drosophila cells: an efficient, homology-dependent pathway of illegitimate recombination near a target site

Drosophila cells in culture can be transformed by introducing exogenous DNA carrying a selectable marker. Here we report on the fate of plasmids that contain an extended fragment of Drosophila DNA in addition to the selectable marker. A small minority of the resulting transformants appear to arise from homologous recombination at the chromosomal target. However, the majority of the insertions are the products of illegitimate events in the vicinity of the target DNA, and they often cause mutations in the targeted region. The efficiency of this process, its homology dependence, and the clustering of the products define a novel transformation pathway that we call "parahomologous targeting."

A model for the mechanism of precise integration of a microinjected transgene

A unique transgenic mouse line has undergone transgene integration in a very precise fashion. The phenotype displayed by mice of the line followed the predicted inheritance patterns for X-linked transgene insertion which has been confirmed. In order to investigate the mechanism of integration the DNA sequence of the transgene and cellular junctions have been determined. A comparison between wild type and transgenic mutant sequences at the site of insertion revealed that there was no loss or rearrangement of cellular DNA upon integration of the transgene. The cellular sequences at the transgene 5' and 3' joins are contiguous in the wild type. The integrant exists as a head to tail tandem dimer with minimal loss of sequence compared with the injected monomer. Analysis of the site of insertion has revealed a 5 bp homology between the 5' end of the transgene and the cellular sequences. In addition, adjacent to the site of insertion within the cellular sequences, there are several sequence motifs implicated in recombination events including a clustering of strong consensus sites of DNA topoisomerase type I and a region of homology to the human minisatellite consensus core sequence, the Escherichia coli Chi site and the meiotic recombination hotspot within the E beta gene of the murine major histocompatibility complex. This clustering of features is likely to have been factorial in the integrity of the insertion event. A model depicting the mechanism of this precise integration is proposed.

Nuclear transfer in the bovine using microinjected donor embryos: assessment of development and deoxyribonucleic acid detection frequency

Bovine embryos that had been microinjected with DNA were examined for their potential use as donor embryos in nuclear transfer. Donor embryos were obtained from oocytes collected by transvaginal oocyte aspiration, matured and fertilized in vitro, microinjected with a murine whey acidic protein-human protein C genomic DNA construct, and cultured in vitro on liver cells of buffalo rat (Rattus norvegicus). Blastomeres from these embryos were transferred into enucleated bovine oocytes received from an abattoir by electrofusion at 40 h postmaturation. Following 7 d of culture, the developmental stage was recorded, and resulting embryos were prepared for analysis by polymerase chain reaction. Embryos that were derived from microinjected donor embryos did not differ from control donor embryos (11 vs. 8.6%) in development to the morula and blastocyst stage. Of the biopsies from 20 microinjected donor embryos, 19 were positive for the injected DNA. Of 37 embryos developing normally, only 12 (32.4%) were positive for the injected DNA. These results indicate that microinjected embryos can be successfully used in a nuclear transfer program to produce additional viable embryos and that these embryos may be reliably screened for the transgene for transfer to recipients.

Frequent deletions and sequence aberrations at the transgene junctions of transgenic mice carrying the papillomavirus regulatory and the SV40 TAg gene sequences

Exogenous DNA microinjected into one-cell mouse zygotes either integrates into the host genome within a short time span, or is rapidly degraded. On integration, a transgene sequence is frequently reiterated. In this report, we describe the enzymatic amplification analysis of transgene junctions of 12 transgenic mice carrying different copy numbers of the same transgene with dissimilar ends. The transgene was composed of the regulatory sequence of the type 18 human papillomavirus linked to the TAg gene of the SV40 virus. Nucleotide sequences of 36 of these junctions were also determined. Deletions were found in 33 (91.7%) of the junctions analysed. At the crossover regions, 55.6% contained short overlapping sequences of one to six nucleotides. Insertions of 2-6 extraneous nucleotides were also found in 8.3% of the transgene junctions. Within a 10-nucleotide sequence on both sides of the transgene junctions, topoisomerase I (topo I) cleavage sites, runs of homogeneous purines or pyrimidiens, alternating purine-pyrimidine tracks and (A-T)-rich sequences were found frequently. Stringent control experiments were also performed to ascertain that the observations made were not artefacts resulting from the polymerase chain reaction. Our data therefore indicate that damage had occurred quite frequently and extensively in our transgene construct. Such transgene damage may also occur to various extents in mice carrying other transgenes. Primary structure of the nucleotide sequences of the injected DNA seems to influence the process of transgene reiteration and aberration.

Transgene detection during early murine embryonic development after pronuclear microinjection

The polymerase chain reaction (PCR) technique was used to detect a whey acidic protein (WAP) gene and transgene presence in mouse ova cultured to various stages of development after pronuclear microinjection at the one-cell stage. The PCR technique detected an endogenous 442 bp WAP DNA sequence in 78% of one-cell, 88% of two-cell and 94% of four-cell ova, and in 95% of morulae and 97% of blastocysts. The heterologous WAP-human protein C transgene was detected in 88% of one-cell, 88% of two-cell and 44% of four-cell ova, and in 40% of morulae and 29% of blastocysts. For comparison, the integration frequency for transgenic mouse production using the same DNA construct was 22%. After five days of in vitro culture, embryos that were either developmentally arrested or fragmented were tested for the presence of the transgene. The injected construct was detected in 83% of arrested one-cell, 85% of arrested two-cell, and 85% of fragmented ova. In culture, only 28% of zygotes microinjected with DNA developed to the blastocysts stage compared to 74% of noninjected zygotes, while 63% of zygotes developed to the blastocyst stage after injection of buffer alone. Pronuclear injection of the transgene at concentrations of 1.5, 15 and 50 micrograms ml-1 resulted in 28, 11 and 9% development to blastocysts and 29, 86 and 88% transgene detection, respectively. Transgene detection was 85, 96 and 97% in degenerate embryos at the respective doses of DNA. These data show that pronuclear microinjection of the transgene is detrimental to subsequent embryonic development.(ABSTRACT TRUNCATED AT 250 WORDS)

Insertional mutagenesis in transgenic mice

Increasing numbers of transgenic mouse lines have resulted in several dozens of mutants created by insertional mutagenesis. The advantages of different vector systems and the problems associated with the analysis of mutations and the cloning of the affected genes are discussed in this review.

Gene transfer efficiency during gestation and the influence of co-transfer of non-manipulated embryos on production of transgenic mice

Litter size of DNA microinjected zygotes is lower than for non-manipulated zygotes. The rate of embryonic and fetal survival in early, mid and late gestation was determined to assess whether DNA integration was responsible for embryonic losses. Also, the effect of including non-microinjected embryos with injected embryos on pregnancy rate and transgenic pup production was determined. In Experiment 1, one-cell embryos from immature CD-1 mice were microinjected with a whey acidic protein promoter-human protein C gene construct. One hour after microinjection embryos were transferred to pseudopregnant recipients (45 transfers of 30 embryos each). Fifteen recipients were sacrificed on day 4, 12 and 18 of gestation and the embryos/fetuses analysed for the transgene. The percentage of embryos or fetuses that were positive for the transgene was not significantly different at any day. However, the number of viable embryos at day 4 was significantly greater than fetuses on days 12 or 18. In addition, a high degree of mosaicism was observed in day 18 fetuses and placentae recovered. In Experiment 2, one-cell embryos from CD-1 mice were microinjected and co-transferred with non-manipulated embryos (C57BL/6). Pregnancy rate and the total number of pups born were improved by addition of non-injected embryos. However, the number of transgenic mice produced was similar whether non-injected embryos were included or not. There were 32.2% (15/46) transgenic pups when 0 non-injected embryos were transferred compared with 15.1% (13/86) transgenic pups when 4 or 8 non-injected embryos were added to the transfers. In summary, a high degree of embryonic and fetal mortality occurs among microinjected embryos.(ABSTRACT TRUNCATED AT 250 WORDS)

Germ-line gene modification and disease prevention: some medical and ethical perspectives

There has been considerable debate about the ethics of human germ-line gene modification. As a result of recent advances in the micromanipulation of embryos and the laboratory development of transgenic mice, a lively discussion has begun concerning both the technical feasibility and the ethical acceptability of human germ-line modification for the prevention of serious disease. This article summarizes some of the recent research on germ-line gene modification in animal models. Certain monogenic deficiency diseases that ultimately might be candidates for correction by germ-line intervention are identified. Several of the most frequently considered ethical issues relative to human germ-line gene modification are considered in the context of professional ethics, parental responsibility, and public policy. Finally, it is suggested that there is merit in continuing the discussion about human germ-line intervention, so that this technique can be carefully compared with alternative strategies for preventing genetic disease.

Gene transfer in bovine blastocysts using replication-defective retroviral vectors packaged with Gibbon ape leukemia virus envelopes

With this work we demonstrate that murine leukemia virus (MLV)-based replication-defective retroviral vectors encapsidated with Gibbon ape leukemia virus (GaLV) envelopes are significantly more infectious to bovine embryonic trachea (EBTr) cells than vectors encapsidated with murine xenotropic envelope proteins. In a test of internal promoter activity in an MLV retroviral vector, the rat beta-actin promoter was shown to be better than the herpes simplex virus type 1 thymidine kinase (TK) and human cytomegalovirus (CMV) immediate early promoters for the expression of an E. coli beta-galactosidase marker gene in bovine target cells. By co-culture of bovine blastocysts and virus-producing cells, or by culture of embryos in the medium harvested from virus-producing cells, we transferred the E. coli beta-galactosidase gene into trophoblasts and also into inner cell mass (ICM) cells of a bovine embryo through the infection of the MLV-based replication-defective retroviruses encapsidated with GaLV envelope proteins. The infection was confirmed by the expression of the E. coli beta-galactosidase gene under a beta-actin internal promoter. In addition, co-culture of ICM cells with virus-producing cells resulted in differentiation of ICM cells into embryoid bodies expressing the marker genes.

A DNA insertional mutation results in microphthalmia in transgenic mice

Transgenic mice were produced by microinjection of a human A gamma-globin gene construct containing site 2 of the locus control region and the A gamma-globin gene with its 3' enhancer sequence. One transgenic mouse line (5'HS2 gamma en91) displayed an altered phenotype when the insertion event of this transgenic line was homozygous. These animals lack the normal pigmentation seen in their hemizygous and non-transgenic littermates, thus appearing white with unpigmented eyes. In addition, their eyes are underdeveloped, consistent with the phenotype associated with mutations at the microphthalmia (mi) locus. Backcrosses of transgenic mice with mi mutant mice result in phenotypes showing a lack of complementation, demonstrating that the site of transgene insertion is allelic with mi. Electron microscopic analysis of hair follicles and culturing of melanocytes from the skin of transgenic animals reveals an absence of cutaneous melanocytes in homozygotes and aberrant growth and morphology of the melanocytes isolated from hemizygous animals. The results presented here summarize the effects of this new allele of the mi locus.

Preferential extrachromosomal localization of exogenous DNA in transgenic silkworm Bombyx mori L

Transgenic silkworms (Bombyx mori L.) were obtained by microinjection of plasmid pPrC-LTR1.5, which carris 1.5 DNA copies of Rous sarcoma virus (RSV) long terminal repeats (LTRs) inserted in the vector pBR322. The transgene was transmitted over the three generations obtained up to now. Most of the exogenous DNA failed to integrate into the genome and persisted as an extrachromosomal element that is subject to rearrangements. Plasmids carrying only part of the input DNA together with fragments of silkworm DNA were rescued from the transgenic animals. One of the rescued plasmids contained a sequence which belongs to a family of evolutionarily conserved repeated sequences.

The majority of G0 transgenic mice are derived from mosaic embryos

Most transgenic mice are generated by the direct microinjection of DNA fragments into the pronuclei of fertilized eggs. It has been generally assumed that the majority of integration events occur prior to the first round of chromosomal DNA replication (Palmiter and Brinster, 1986). In this study we have determined by comparison of PCR, Southern blot and transmission frequencies that at least 62% of integration events generate a mosaic (somatic and/or germline) G0 transgenic mouse. Furthermore, the statistical probability of transgene-containing cells segregating to the various early embryo lineages implies that this is probably an underestimate of the true mosaic frequency. Thus, the majority of DNA injected into fertilized mouse eggs intergates after the first round of chromosomal DNA replication, therefore most G0 transgenic mice are derived from a mosaic embryo.

Inherited haemolytic anaemia created by insertional inactivation of the alpha-spectrin gene

In the process of generating transgenic mice, inserted foreign DNA can cause insertional inactivation of the flanking genetic locus and simultaneously provide a molecular tag for localizing and cloning the inactivated gene. We describe the case of an insertional mutation leading, in animals homozygous for the insertion, to severe anaemia that was lethal within a few days after birth. The haemolytic anaemia and microspherocytosis of the red cells strongly suggested membrane abnormalities of the erythrocytes. By in situ localization of the integration site, protein analysis of the red cell membranes, northern and Southern blot analyses, we were able to demonstrate that the integrated transgene had affected the alpha-spectrin gene locus.

Structure, expression, and chromosome location of the gene for the beta subunit of brain-specific Ca2+/calmodulin-dependent protein kinase II identified by transgene integration in an embryonic lethal mouse mutant

The transgenic mouse strain CAT40 carries in its germ line one copy of a DNA construct consisting of the chloramphenicol acetyltransferase gene and the immunoglobulin heavy-chain enhancer. We show that transgene integration has resulted in a recessive lethal mutation that leads to death of homozygous CAT40 embryos shortly after implantation. The transgene has integrated adjacent to the 3' end of the gene coding for the beta subunit of the brain-specific Ca2+/calmodulin-dependent protein kinase II (Camk-2). The complete cDNA sequence of the Camk-2 gene and most of its exon/intron structure was determined. The deduced amino acid sequence is highly homologous to the previously described rat protein. The chromosomal location of the Camk-2 locus was mapped by interspecific backcross analysis to the proximal region of mouse chromosome 11. This region lacks previously identified recessive embryonic lethal mutations. During embryonic development, Camk-2-specific transcripts are first seen in the head section of 12.5-day-old embryos, and in adult mice the gene is expressed almost exclusively in the brain. Although transcription of the Camk-2 gene in heterozygous CAT40 mice is affected by transgene integration, it is unlikely that this gene is responsible for the mutant phenotype, since it is not expressed in blastocysts and the first transcripts during normal development are detected after the death of homozygous CAT40 embryos. Transgene integration is accompanied by a large deletion of cellular DNA; death is therefore most likely caused by the loss of a gene or genes that are important for early postimplantation development.

Structure, expression, and chromosome location of the gene for the beta subunit of brain-specific Ca2+/calmodulin-dependent protein kinase II identified by transgene integration in an embryonic lethal mouse mutant

The transgenic mouse strain CAT40 carries in its germ line one copy of a DNA construct consisting of the chloramphenicol acetyltransferase gene and the immunoglobulin heavy-chain enhancer. We show that transgene integration has resulted in a recessive lethal mutation that leads to death of homozygous CAT40 embryos shortly after implantation. The transgene has integrated adjacent to the 3' end of the gene coding for the beta subunit of the brain-specific Ca2+/calmodulin-dependent protein kinase II (Camk-2). The complete cDNA sequence of the Camk-2 gene and most of its exon/intron structure was determined. The deduced amino acid sequence is highly homologous to the previously described rat protein. The chromosomal location of the Camk-2 locus was mapped by interspecific backcross analysis to the proximal region of mouse chromosome 11. This region lacks previously identified recessive embryonic lethal mutations. During embryonic development, Camk-2-specific transcripts are first seen in the head section of 12.5-day-old embryos, and in adult mice the gene is expressed almost exclusively in the brain. Although transcription of the Camk-2 gene in heterozygous CAT40 mice is affected by transgene integration, it is unlikely that this gene is responsible for the mutant phenotype, since it is not expressed in blastocysts and the first transcripts during normal development are detected after the death of homozygous CAT40 embryos. Transgene integration is accompanied by a large deletion of cellular DNA; death is therefore most likely caused by the loss of a gene or genes that are important for early postimplantation development.

Increased postimplantation loss and malformations among the F2 progeny of male rats chronically treated with cyclophosphamide

Cyclophosphamide, administered to the male rat, produces increased pre- and postimplantation loss in the progeny as well as an increase in the numbers of malformed and growth retarded fetuses. The purpose of this study was to determine whether the adverse effects of chronic paternal cyclophosphamide exposure are transmissible to the next generation, the F2 progeny. Adult male rats were treated by gavage daily with saline or with cyclophosphamide (3.4 or 5.1 mg/kg) for 4 or 18 weeks and mated. The male and female offspring in each treatment group (F1 generation) were randomly mated. The resulting pregnant females were killed on day 20 of gestation to evaluate progeny outcome in the F2 generation. There was a significant increase in postimplantation loss among the offspring of the group whose fathers had been treated with cyclophosphamide at a dose of 5.1 mg/kg/day. Exposure to a dose of 5.1 mg/kg/day of cyclophosphamide also resulted in an F2 generation with a significantly decreased mean fetal weight per litter and a significant increase in the number of malformed fetuses. The malformations observed among the F2 progeny included open eyes, omphalocele, generalized edema, syndactyly, gigantism, and dwarfism. Thus, exposure of the father to cyclophosphamide does result in a specific and heritable alteration in the fertility of the surviving "apparently normal" F1 progeny. Interestingly, the adverse consequences of exposure of male rats to cyclophosphamide are similar in the F2 generation to those previously reported for the F1 progeny.

Paternal cyclophosphamide treatment causes postimplantation loss via inner cell mass-specific cell death

Treatment of the father with the anticancer alkylating agent cyclophosphamide has negative effects on embryonic development in the rat. Four-week treatment of male rats with a low dose of cyclophosphamide causes a dramatic, dose-dependent increase in postimplantation death of the progeny. Several recent studies have indicated that the paternal genome is required for the development of the extraembryonic tissues. Thus, the purpose of this study was to determine which tissues of the implanting embryo were affected by paternal exposure to cyclophosphamide. Male Sprague-Dawley rats were given cyclophosphamide (6 mg/kg/day) or saline by gavage and bred to untreated female rats after 4 weeks of treatment. Pregnant female rats were killed on day 7 of gestation, and implantation sites were dissected from the uterus, fixed, embedded in Epon for semithin serial sectioning, and stained for subsequent light microscopy. Strikingly, many of the implantation sites of affected embryos sired by treated males displayed an apparently normal trophectoderm enclosing a region of dying cells, containing dark-stained pyknotic nuclei. Very few or no inner cell mass-derived embryonic cells were present in these implantation sites. Therefore, there is a selective death of inner cell mass-derived cells in day 7 implantation sites obtained from the progeny of cyclophosphamide-treated males. The results of this study suggest that treatment of the male with cyclophosphamide can affect paternal genes specifically required for development of the inner cell mass cells of the embryo, without an apparent effect on those genes required for normal trophectoderm.

Induced reciprocal translocation in transgenic mice near sites of transgene integration

Transgenic mice (JCP0 #18), heterozygous for an insertion of approximately 50 copies of the rat peripheral myelin (P0) protein cDNA, displayed a pattern of reduced litter size that suggested a chromosome rearrangement. Chromosome banding studies of fetal cells disclosed the presence of an apparently balanced translocation between a Chromosome (Chr) 1 and 14 with breakpoints at bands 1H3 and 14C3. In situ hybridization of biotin-labeled P0 rat cDNA probe to chromosome spreads and detection of specific signal with fluorescein isothiocyanate-conjugated avidin revealed a strong signal on the 1(14) translocation chromosome at the site of the breakpoint. A weaker signal was present near the breakpoint on the 14(1) derivative chromosome. These results suggest an etiologic relationship between the insertion of the transgene and the origin of the translocation. To further elucidate possible mechanisms, we first mapped the endogenous P0 gene (gene symbol Mpp). As previously reported (You et al., Genomics 9: 751, 1991), we found that Mpp is located on Chr 1 in the region of the translocation breakpoint in JCP0 mice. Subsequently, we have carried out pulsed-field gel and standard Southern analyses with P0 gene probes, but found no evidence for a direct involvement of the endogenous P0 gene in the process that generated the balanced reciprocal translocation. Thus, we favor the hypothesis that, during repair of DNA strand breakage--possibly induced by the microinjection procedure--the transgene copies were ligated to broken ends of Chrs 1 and 14. According to convention, this translocation is designated T(1;14)1Po. Homozygotes are phenotypically normal and breed well; they will be useful for genetic and physical mapping of Chrs 1 and 14.

The same genomic region is disrupted in two transgene-induced limb deformity alleles

Mutations of the mouse limb deformity locus, ld, map to Chromosome (Chr) 2 and result in defects in the morphogenesis and patterning of the limb and kidney. Complementation studies have defined the existence of five recessive ld alleles. Remarkably, two of these, ldTgHd and ldTgBri, are transgene-induced mutations. Recovery of the first transgene insertional allele, ldTgHd, facilitated the molecular cloning of a large (greater than 200 kb) candidate gene at the ld locus. This gene is broadly transcribed and encodes a set of novel protein isoforms, termed formins. Here we present characterization of the ldTgBri mutation that supports the molecular identification of the ld gene. We show that the ldTgBri fails to complement both the ldTgHd and the ldOR alleles and that it has undergone a genomic deletion that disrupts the cloned ld gene and its transcripts. Curiously, the ldTgBri deletion encompasses the same 11-kb interval in which the ldTgHd insertion occurred and in which a chromosomal rearrangement has been identified in a third allele, ldIn2. These findings suggest that this region of the ld gene is a preferential site for illegitimate recombination.

Mosaic expression of a tyrosinase fusion gene in albino mice yields a heritable striped coat color pattern in transgenic homozygotes

Genetically albino mouse eggs were injected with an inducible transgene comprising the wild-type tyrosinase (monophenol, L-dopa: oxygen oxidoreductase, EC 1.14.18.1) cDNA and the metallothionein gene promoter in the expectation that the transgene would be expressed to different extents in the various developing pigment cell clones of at least some individuals, thereby leading to patterned coats. This proved to be the case. Five transgenic mice had lightly pigmented patterns of transverse stripes visualizing melanoblast proliferation and migration dorsoventrally on each side. Similar patterns have been seen in genetically mosaic mouse models produced from conjoined blastomeres of different color genotypes and in many naturally patterned genotypes of mice. Four of the transgenics had heritable patterns and autosomal transgene integration. Their homozygous descendants were darker than hemizygotes and transmitted the basic pattern through many generations. Eyes were also pigmented, with clonal patches of melanized cells in the retinal pigment epithelium. The skin was dark due to many pigmented dermal melanocytes, whereas relatively few were in the hair follicles. This "inversion" is attributable to precocious maturation and migratory arrest of many melanoblasts during passage through the dermis en route to the hair bulbs. Patterning in these mice is considered in light of the view, previously proposed, that phenotypically different clones, or phenoclones, may exist in virtually all mammalian cell types and that many genes may be associated with cis-acting control regions causing variations in their expression that are mitotically perpetuated. We point out that mosaic expression has many implications for development as well as neoplasia. In the latter case, the potential for tumor susceptibility may be affected by clonal variation without further gene mutations or deletions. Thus, mice with variegating transgenes can provide molecular access to gene control mechanisms and to their consequences in development and disease.

Molecular basis of mouse developmental mutants

Collapse

Tg (9 HSA-MYC), a homozygous lethal insertion in the mouse

Transgenic mice generated with different DNA sequences were surveyed for possible homozygous mutant phenotypes. We found an embryonic lethal mutation in the transgenic mouse strain (MT-MYC12.4) containing the human c-myc gene. Embryos homozygous for the transgene die shortly after implantation. The strain MT-MYC12.4 carries approximately 50 tandem copies of the recombinant plasmid sequence. The 3' flanking sequence has been cloned and analyzed. It contains a unique sequence that has been conserved during evolution and maps to Chromosome (Chr) 9. This mutant has been designated Tg 9 (HSA-MYC).

Malignant melanoma in transgenic mice

Ocular and cutaneous melanomas arose in new inbred lines of transgenic mice having an integrated recombinant gene comprised of the tyrosinase promoter, expressed in pigment cells, and the simian virus 40 early-region transforming sequences. The tumors were hypomelanotic and were histopathologically similar to corresponding human melanomas. Eye melanomas often originated at a young age, chiefly from the retinal pigment epithelium, also from the choroid, and rarely from the ciliary body. The eye tumors grew aggressively, were highly invasive, and metastasized to local and distant sites. The earliest formation of these tumors was associated with higher copy numbers of the transgene; mice of different single-copy lines varied greatly in age of onset and frequency of eye tumors. Coat pigmentation was reduced in almost all lines, to various extents. Primary skin melanomas arose later and less frequently than eye melanomas. Hence they were at early stages and of unknown long-range incidence in this investigation, in which autopsies covered the first half-year of life. For both ocular and cutaneous melanomas, the transgenic mice offer numerous possibilities for experimental study of mechanisms underlying formation and spread of melanomas.

Direct sequencing of PCR-amplified junction fragments from tandemly repeated transgenes

When microinjected foreign genes integrate into the genomes of mice, multiple copies are frequently found clustered together at one location. How they concatamerize--by the integration of large linearized concatamers that are formed by simple end-to-end linkage, by circularization of individual DNA fragments and recombination, or by some other means--is not understood. In the transgenic animals studied thus far by ourselves and others, integration frequency and transgene copy number do not seem to be significantly influenced by the complementarity of the ends of the DNA fragments that have been microinjected. We have utilized PCR amplification and DNA sequence analysis to study selected transgene junctions at the nucleotide level. In two transgenic mice carrying the synthetic RSVcat gene (injected with noncomplementary overhangs on the fragment ends), ends were 'nibbled' from 1 to 62 bases before being joined to an adjacent gene copy. Repeated dinucleotides, providing the most minimal of homologies, are present in half of the characterized junctions. Determination of the relative copy number of the junctions in each mouse supports the idea that transgene complexes can undergo additional rearrangements after the initial formation event.

Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants

Plasmids comprising transgene insertions in four lines of transgenic mice have been retrieved by plasmid rescue into a set of Escherichia coli strains with mutations in different members of the methylation-dependent restriction system (MDRS). Statistical analysis of plasmid rescue frequencies has revealed that the MDRS loci detect differential modifications of the transgene insertions among mouse lines that show distinctive patterns of transgene expression. Plasmids in mice that express hybrid insulin transgenes during development can be readily cloned into E. coli strains carrying mutations in two of the MDRS loci, mcrA and mcrB. In mice in which transgene expression is inappropriately delayed into adulthood, plasmids can only be cloned into E. coli that carry mutations in all known MDRS activities. Differential cloning frequencies in the presence or absence of the various methylation-dependent restriction genes represent a further way to distinguish regions of mammalian chromosomes. These multiply deficient E. coli strains will also facilitate the molecular cloning of modified chromosomal DNA.

Expression of a foreign gene in a line of transgenic mice is modulated by a chromosomal position effect

Unusual aberrant expression of a foreign gene in a particular transgenic mouse line is often attributed to chromosomal position effect, although proof of this is lacking. An alternative explanation is that expression has been modified by the arrangement of multiple copies of the foreign gene at the insertion site or by mutation or gene rearrangement. We have distinguished between these explanations in the case of one particular transgenic line by recovering the aberrantly expressed foreign DNA and reintroducing it into the mouse genome to produce secondary transgenic mice. The expression pattern of the gene in the secondary transgenic mice was normal, showing that this case of aberrant expression is due to a chromosomal position effect.

Expression of a foreign gene in a line of transgenic mice is modulated by a chromosomal position effect

Unusual aberrant expression of a foreign gene in a particular transgenic mouse line is often attributed to chromosomal position effect, although proof of this is lacking. An alternative explanation is that expression has been modified by the arrangement of multiple copies of the foreign gene at the insertion site or by mutation or gene rearrangement. We have distinguished between these explanations in the case of one particular transgenic line by recovering the aberrantly expressed foreign DNA and reintroducing it into the mouse genome to produce secondary transgenic mice. The expression pattern of the gene in the secondary transgenic mice was normal, showing that this case of aberrant expression is due to a chromosomal position effect.