Testing an "in-out" targeting procedure for making subtle genomic modifications in mouse embryonic stem cells

Seamless Gene Correction in the Human Cystic Fibrosis Transmembrane Conductance Regulator Locus by Vector Replacement and Vector Insertion Events

Background: Gene correction via homology directed repair (HDR) in patient-derived induced pluripotent stem (iPS) cells for regenerative medicine are becoming a more realistic approach to develop personalized and mutation-specific therapeutic strategies due to current developments in gene editing and iPSC technology. Cystic fibrosis (CF) is the most common inherited disease in the Caucasian population, caused by mutations in the CF transmembrane conductance regulator (CFTR) gene. Since CF causes significant multi-organ damage and with over 2,000 reported CFTR mutations, CF patients could be one prominent population benefiting from gene and cell therapies. When considering gene-editing techniques for clinical applications, seamless gene corrections of the responsible mutations, restoring native "wildtype" DNA sequence without remnants of drug selectable markers or unwanted DNA sequence changes, would be the most desirable approach. Result: The studies reported here describe the seamless correction of the W1282X CFTR mutation using CRISPR/Cas9 nickases (Cas9n) in iPS cells derived from a CF patient homozygous for the W1282X Class I CFTR mutation. In addition to the expected HDR vector replacement product, we discovered another class of HDR products resulting from vector insertion events that created partial duplications of the CFTR exon 23 region. These vector insertion events were removed via intrachromosomal homologous recombination (IHR) enhanced by double nicking with CRISPR/Cas9n which resulted in the seamless correction of CFTR exon 23 in CF-iPS cells. Conclusion: We show here the removal of the drug resistance cassette and generation of seamless gene corrected cell lines by two independent processes: by treatment with the PiggyBac (PB) transposase in vector replacements or by IHR between the tandemly duplicated CFTR gene sequences.

To CRISPR and beyond: the evolution of genome editing in stem cells

The goal of editing the genomes of stem cells to generate model organisms and cell lines for genetic and biological studies has been pursued for decades. There is also exciting potential for future clinical impact in humans. While recent, rapid advances in targeted nuclease technologies have led to unprecedented accessibility and ease of gene editing, biology has benefited from past directed gene modification via homologous recombination, gene traps and other transgenic methodologies. Here we review the history of genome editing in stem cells (including via zinc finger nucleases, transcription activator-like effector nucleases and CRISPR-Cas9), discuss recent developments leading to the implementation of stem cell gene therapies in clinical trials and consider the prospects for future advances in this rapidly evolving field.

Influence of threonine metabolism on S-adenosylmethionine and histone methylation

Threonine is the only amino acid critically required for the pluripotency of mouse embryonic stem cells (mESCs), but the detailed mechanism remains unclear. We found that threonine and S-adenosylmethionine (SAM) metabolism are coupled in pluripotent stem cells, resulting in regulation of histone methylation. Isotope labeling of mESCs revealed that threonine provides a substantial fraction of both the cellular glycine and the acetyl-coenzyme A (CoA) needed for SAM synthesis. Depletion of threonine from the culture medium or threonine dehydrogenase (Tdh) from mESCs decreased accumulation of SAM and decreased trimethylation of histone H3 lysine 4 (H3K4me3), leading to slowed growth and increased differentiation. Thus, abundance of SAM appears to influence H3K4me3, providing a possible mechanism by which modulation of a metabolic pathway might influence stem cell fate.

Cardiac-specific inducible and conditional gene targeting in mice

Mouse genetic engineering has revolutionized our understanding of the molecular and genetic basis of heart development and disease. This technology involves conditional tissue-specific and temporal transgenic and gene targeting approaches, as well as introduction of polymorphisms into the mouse genome. These approaches are increasingly used to elucidate the genetic pathways underlying tissue homeostasis, physiology, and pathophysiology of adult heart. They have also led to the development of clinically relevant models of human cardiac diseases. Here, we review the technologies and their limitations in general and the cardiovascular research community in particular.

Efficient gene targeting by homologous recombination in rat embryonic stem cells

The rat is the preferred experimental animal in many biological studies. With the recent derivation of authentic rat embryonic stem (ES) cells it is now feasible to apply state-of-the art genetic engineering in this species using homologous recombination. To establish whether rat ES cells are amenable to in vivo recombination, we tested targeted disruption of the hypoxanthine phosphoribosyltransferase (hprt) locus in ES cells derived from both inbred and outbred strains of rats. Targeting vectors that replace exons 7 and 8 of the hprt gene with neomycinR/thymidine kinase selection cassettes were electroporated into male Fisher F344 and Sprague Dawley rat ES cells. Approximately 2% of the G418 resistant colonies also tolerated selection with 6-thioguanine, indicating inactivation of the hprt gene. PCR and Southern blot analysis confirmed correct site-specific targeting of the hprt locus in these clones. Embryoid body and monolayer differentiation of targeted cell lines established that they retained differentiation potential following targeting and selection. This report demonstrates that gene modification via homologous recombination in rat ES cells is efficient, and should facilitate implementation of targeted, genetic manipulation in the rat.

Overview: generation of gene knockout mice

The technique of gene targeting allows for the introduction of engineered genetic mutations into a mouse at a determined genomic locus. The process of generating mouse models with targeted mutations was developed through both the discovery of homologous recombination and the isolation of murine embryonic stem cells (ES cells). Homologous recombination is a DNA repair mechanism that is employed in gene targeting to insert a designed mutation into the homologous genetic locus. Targeted homologous recombination can be performed in murine ES cells through electroporation of a targeting construct. These ES cells are totipotent and, when injected into a mouse blastocyst, they can differentiate into all cell types of a chimeric mouse. A chimeric mouse harboring cells derived from the targeted ES cell clone can then generate a whole mouse containing the desired targeted mutation. The initial step for the generation of a mouse with a targeted mutation is the construction of an efficient targeting vector that will be introduced into the ES cells.

Genetic approaches for changing the heart and dissecting complex syndromes

The genetic, biochemical and molecular bases of human cardiac disease have been the focus of extensive research efforts for many years. Early animal models of cardiovascular disease used pharmacologic or surgical interventions, or took advantage of naturally occurring genetic abnormalities and the data obtained were largely correlative. The inability to directly alter an organism's genetic makeup and cellular protein content and accurately measure the results of that manipulation precluded rigorous examination of true cause-effect and structure-function relationships. Directed genetic manipulation in the mouse gave researchers the ability to modify and control the mammalian heart's protein content, resulting in the rational design of models that could provide critical links between the mutated or absent protein and disease. Two techniques that have proven particularly useful are transgenesis, which involves the random insertion of ectopic genetic material of interest into a "host" genome, and gene targeting, which utilizes homologous recombination at a pre-selected locus. Initially, transgenesis and gene targeting were used to examine systemic loss-of-function and gain-of-function, respectively, but further refinements in both techniques have allowed for investigations of organ-specific, cell type-specific, developmental stage-sensitive and dose-dependent effects. Genetically engineered animal models of pediatric and adult cardiac disease have proven that, when used appropriately, these tools have the power to extend mere observation to the establishment of true causative proof. We illustrate the power of the general approach by showing how genetically engineered mouse models can define the precise signaling pathways that are affected by the gain-of-function mutation that underlies Noonan syndrome. Increasingly precise and modifiable animal models of human cardiac disease will allow researchers to determine not only pathogenesis, but also guide treatment and the development of novel therapies.

Remodeling the cardiac sarcomere using transgenesis

An underpinning of basic physiology and clinical medicine is that specific protein complements underlie cell and organ function. In the heart, contractile protein changes correlating with functional alterations occur during both normal development and the development of numerous pathologies. What has been lacking for the majority of these observations is an extension of correlation to causative proof. More specifically, different congenital heart diseases are characterized by shifts in the motor proteins, and the genetic etiologies of a number of different dilated and hypertrophic cardiomyopathies have been established as residing at loci encoding the contractile proteins. To establish cause, or to understand development of the pathophysiology over an animal's life span, it is necessary to direct the heart to synthesize, in the absence of other pleiotropic changes, the candidate protein. Subsequently one can determine whether or how the protein's presence causes the effects either directly or indirectly. By affecting the heart's protein complement in a defined manner, the potential to establish the function of different proteins and protein isoforms exists. Transgenesis provides a means of stably modifying the mammalian genome. By directing expression of engineered proteins to the heart, cardiac contractile protein profiles can be effectively remodeled and the resultant animal used to study the consequences of a single, genetic manipulation at the molecular, biochemical, cytological, and physiological levels.

Enhancing the efficiency of introducing precise mutations into the mouse genome by hit and run gene targeting

The creation of precise clinical mutations by targeting is important in elucidating disease pathogenesis using mouse models. 'Hit and run' gene targeting is an elegant method to achieve this goal. This uses first a positive selection to introduce the targeting vector carrying the required mutation and then a negative selection to identify clones which have removed vector and wild-type sequences by intrachromosomal recombination. However, this approach has only been successfully used in a handful of cases. We used this procedure to introduce precise clinical mutations into the exon 10 region of the cystic fibrosis transmembrane conductance regulator (Cftr) gene. Using a CMV promoter driven hygromycin/thymidine kinase (hyg/tk) fusion gene as both our dominant and negative selectable marker, we targeted the Cftr locus very efficiently but only identified false runs after the negative selection step. This defect in thymidine kinase induced toxicity to gancyclovir correlated with methylation of the transgene. Consequently we devised a stringent screening procedure to select only true 'run' clones. Unfortunately these 'run' clones had lost the mutation so we altered the vector design to bias the run step to retain the mutation and used a different tk selection cassette with a HSVtk promoter sequence. This new vector design allowed both efficient 'hit and run' for two cystic fibrosis (CF) mutations with no false positives and successful germline transmission of the novel G480C missense mutation.

Subtle mutagenesis by ends-in recombination in malaria parasites

The recent advent of gene-targeting techniques in malaria (Plasmodium) parasites provides the means for introducing subtle mutations into their genome. Here, we used the TRAP gene of Plasmodium berghei as a target to test whether an ends-in strategy, i.e., targeting plasmids of the insertion type, may be suitable for subtle mutagenesis. We analyzed the recombinant loci generated by insertion of linear plasmids containing either base-pair substitutions, insertions, or deletions in their targeting sequence. We show that plasmid integration occurs via a double-strand gap repair mechanism. Although sequence heterologies located close (less than 450 bp) to the initial double-strand break (DSB) were often lost during plasmid integration, mutations located 600 bp and farther from the DSB were frequently maintained in the recombinant loci. The short lengths of gene conversion tracts associated with plasmid integration into TRAP suggests that an ends-in strategy may be widely applicable to modify plasmodial genes and perform structure-function analyses of their important products.

Introduction of hereditary disease-associated mutations into the beta-amyloid precursor protein gene of mouse embryonic stem cells: a comparison of homologous recombination methods

Two different approaches for introducing pathogenic mutations into the beta-amyloid precursor protein gene in mouse embryonic stem cells were compared. Both approaches require two sequential modifications of the targeting locus by homologous recombinations. One approach was a "targeting-in-out" procedure that is based on a double-replacement strategy, and the other was a "hit-and-run" procedure that makes use of an unstable genomic duplication after vector integration. Both approaches showed similar targeting frequencies for the first step. In the targeting-in-out procedure, targeted-in embryonic stem cell clones with the desired mutation and an intron-located selection cassette were obtained at a high frequency after the first step. Targeting out, however, resulted not only in the expected loss of the intron-located selection cassette but also in unavoidable reversion to wild type. In contrast, pure mutants, i.e., those without additional genomic changes, were generated by the hit-and-run procedure. Although targeted-in embryonic stem cells might be used to generate animals with modified beta-amyloid precursor protein, the hit-and-run procedure appears to be the superior way to target gene modifications in vivo, leading to pure, correct mutants. For further improvements, optimization of the homologous recombination efficiency could be envisaged.

Analysis of gene targeting and intrachromosomal homologous recombination stimulated by genomic double-strand breaks in mouse embryonic stem cells

To investigate the effects of in vivo genomic DNA double-strand breaks on the efficiency and mechanisms of gene targeting in mouse embryonic stem cells, we have used a series of insertion and replacement vectors carrying two, one, or no genomic sites for the rare-cutting endonuclease I-SceI. These vectors were introduced into the hypoxanthine phosphoribosyltransferase (hprt) gene to produce substrates for gene-targeting (plasmid-to-chromosome) or intrachromosomal (direct repeat) homologous recombination. Recombination at the hprt locus is markedly increased following transfection with an I-SceI expression plasmid and a homologous donor plasmid (if needed). The frequency of gene targeting in clones with an I-SceI site attains a value of 1%, 5,000-fold higher than that in clones with no I-SceI site. The use of silent restriction site polymorphisms indicates that the frequencies with which donor plasmid sequences replace the target chromosomal sequences decrease with distance from the genomic break site. The frequency of intrachromosomal recombination reaches a value of 3.1%, 120-fold higher than background spontaneous recombination. Because palindromic insertions were used as polymorphic markers, a significant number of recombinants exhibit distinct genotypic sectoring among daughter cells from a single clone, suggesting the existence of heteroduplex DNA in the original recombination product.

A gene-targeting approach identifies a function for the first intron in expression of the alpha1(I) collagen gene

The role of the first intron of the Col1A1 gene in the regulation of type I collagen synthesis remains uncertain and controversial despite numerous studies that have made use of transgenic and transfection experiments. To examine the importance of the first intron in regulation of the gene, we have used the double-replacement method of gene targeting to introduce, by homologous recombination in embryonic stem (ES) cells, a mutated Col1A1 allele (Col-IntDelta). The Col-IntDelta allele contains a 1. 3-kb deletion within intron I and is also marked by the introduction of a silent mutation that created an XhoI restriction site in exon 7. Targeted mice were generated from two independently derived ES cell clones. Mice carrying two copies of the mutated gene were born in the expected Mendelian ratio, developed normally, and showed no apparent abnormalities. We used heterozygous mice to determine whether expression of the mutated allele differs from that of the normal allele. For this purpose, we developed a reverse transcription-PCR assay which takes advantage of the XhoI polymorphism in exon 7. Our results indicate that in the skin, and in cultured cells derived from the skin, the intron plays little or no role in constitutive expression of collagen I. However, in the lungs of young mice, the mutated allele was expressed at about 75% of the level of the normal allele, and in the adult lung expression was decreased to less than 50%. These results were confirmed by RNase protection assays which demonstrated a two- to threefold decrease in Col1A1 mRNA in lungs of homozygous mutant mice. Surprisingly, in cultured cells derived from the lung, the mutated allele was expressed at a level similar to that of the wild-type allele. Our results also indicated an age-dependent requirement for the intact intron in expression of the Col1A1 gene in muscle. Since the intron is spliced normally, and since the mutant allele is expressed as well as the wild-type allele in the skin, reduced mRNA stability is unlikely to contribute to the reduction in transcript levels. We conclude that the first intron of the Col1A1 gene plays a tissue-specific and developmentally regulated role in transcriptional regulation of the gene. Our experiments demonstrate the utility of gene-targeting techniques that produce subtle mutations for studies of cis-acting elements in gene regulation.

Factors affecting the efficiency of introducing precise genetic changes in ES cells by homologous recombination: tag-and-exchange versus the Cre-loxp system

The introduction of genetic modifications in specific genes by homologous recombination provides a powerful tool for elucidation of structure-function relationships of proteins of biological interest. Presently, there are several alternative methods of homologous recombination that permit the introduction of small genetic modifications in specific loci. Two of the most widely used methods are the tag-and-exchange, based on the use of positive-negative selection markers, and the Cre-loxP system, based on the use of a site-specific recombinase. The efficiency of detection of targeting events at different loci using the two systems was compared. Additionally, we analysed how the distance between two gene markers placed within the region of homology of a targeting vector affects the rate at which both markers are introduced into the locus during the homologous recombination event. Our results indicate that the method based on the use of positive-negative selection markers was less efficient than the Cre-loxP based system, irrespective of locus or type of positive-negative selection. It was also determined that as the distance between the selectable marker and the genetic modification being introduced increases, there is a progressive reduction in the efficiency of detecting events with the desired genetic modification.

Mice with type 2 and 3 Gaucher disease point mutations generated by a single insertion mutagenesis procedure

Gaucher disease is caused by mutations in the gene encoding the lysosomal enzyme glucocerebrosidase (GC). Three clinical types of Gaucher disease have been defined according to the presence (type 2 and 3) or absence (type 1) of central nervous system disease and severity of clinical manifestations. The clinical course of the disease correlates with the mutation carried by the GC gene. To produce mice with point mutations that correspond to the clinical types of Gaucher disease, we have devised a highly efficient one-step mutagenesis method-the single insertion mutagenesis procedure (SIMP)-to introduce human disease mutations into the mouse GC gene. By using SIMP, mice were generated carrying either the very severe RecNciI mutation that can cause type 2 disease or the less severe L444P mutation associated with type 3 disease. Mice homozygous for the RecNciI mutation had little GC enzyme activity and accumulated glucosylceramide in brain and liver. In contrast, the mice homozygous for the L444P mutation had higher levels of GC activity and no detectable accumulation of glucosylceramide in brain and liver. Surprisingly, both point mutation mice died within 48 hr of birth, apparently of a compromised epidermal permeability barrier caused by defective glucosylceramide metabolism in the epidermis.

Parameters controlling the rate of gene targeting frequency in the protozoan parasite Leishmania

In this study we investigated the role of several parameters governing the efficiency of gene targeting mediated by homologous recombination in the protozoan parasite Leishmania. We evaluated the relative targeting frequencies of different replacement vectors designed to target several sequences within the parasite genome. We found that a decrease in the length of homologous sequences <1 kb on one arm of the vector linearly influences the targeting frequency. No homologous recombination was detected, however, when the flanking homologous regions were <180 bp. A requirement for a very high degree of homology between donor and target sequences was found necessary for efficient gene targeting in Leishmania , as targeted recombination was strongly affected by base pair mismatches. Targeting frequency increased proportionally with copy number of the target only when the target was part of a linear amplicon, but remained unchanged when it was present on circles. Different chromosomal locations were found to be targeted with significantly variable levels of efficiency. Finally, different strains of the same species showed differences in gene targeting frequency. Overall, gene targeting mediated by homologous recombination in Leishmania shares similarities to both the yeast and the mammalian recombination systems.

Substitution of a mutant alpha2a-adrenergic receptor via "hit and run" gene targeting reveals the role of this subtype in sedative, analgesic, and anesthetic-sparing responses in vivo

Norepinephrine contributes to antinociceptive, sedative, and sympatholytic responses in vivo, and alpha2 adrenergic receptor (alpha2AR) agonists are used clinically to mimic these effects. Lack of subtype-specific agonists has prevented elucidation of the role that each alpha2AR subtype (alpha2A, alpha2B, and alpha2C) plays in these central effects. Here we demonstrate that alpha2AR agonist-elicited sedative, anesthetic-sparing, and analgesic responses are lost in a mouse line expressing a subtly mutated alpha2AAR, D79N alpha2AAR, created by two-step homologous recombination. These functional changes are accompanied by failure of the D79N alpha2AAR to inhibit voltage-gated Ca2+ currents and spontaneous neuronal firing, a measure of K+ current activation. These results provide definitive evidence that the alpha2AAR subtype is the primary mediator of clinically important central actions of alpha2AR agonists and suggest that the D79N alpha2AAR mouse may serve as a model for exploring other possible alpha2AAR functions in vivo.

Segmental genomic replacement by Cre-mediated recombination: genotoxic stress activation of the p53 promoter in single-copy transformants

Genotoxic stress results in transcriptional activation of the p53 promoter. To gain more detailed information on genotoxic induction of the p53 promoter at a uniform genomic locus, we have developed an efficient strategy for replacing a defined genomic segment in mouse NIH 3T3 cells with exogenous transfected DNA using a 'double lox' targeting strategy mediated by Cre DNA recombinase. The strategy utilizes a pair of heterospecific lox sites engineered both into the genome and onto the targeting DNA. This allows direct replacement of genomic DNA by a Cre-catalyzed double crossover event. p53-CAT reporter constructs were site-specifically placed into the genomic target 20-fold more efficiently by double lox recombination than by Cre-mediated single crossover insertional recombination, and the absolute frequency of site-specific double lox targeting exceeded the frequency of transformation due to random illegitimate recombination of transfected DNA into the genome. Resulting targeted single-copy integrants of the p53-CAT reporter show strong genotoxic induction by mitomycin C, and a dynamic range of induction that exceeds that seen in transient transfection assays. The double lox strategy is generally applicable to Cre-mediated genomic targeting in any cell and should be of particular utility in the site-specific targeting of DNA into embryonic stem (ES) cells for the production of gene-modified mice.

Phenotypic analysis of mice expressing exclusively apolipoprotein B48 or apolipoprotein B100

Apolipoprotein (apo)-B is found in two forms in mammals: apo-B100, which is made in the liver and the yolk sac, and apo-B48, a truncated protein made in the intestine. To provide models for understanding the physiologic purpose for the two forms of apo-B, we used targeted mutagenesis of the apo-B gene to generate mice that synthesize exclusively apo-B48 (apo-B48-only mice) and mice that synthesize exclusively apo-B100 (apo-B100-only mice). Both the apo-B48-only mice and apo-B100-only mice developed normally, were healthy, and were fertile. Thus, apo-B48 synthesis was sufficient for normal embryonic development, and the synthesis of apo-B100 in the intestines of adult mice caused no readily apparent adverse effects on intestinal function or nutrition. Compared with wild-type mice fed a chow diet, the levels of low density lipoprotein (LDL)-cholesterol and very low density lipoprotein- and LDL-triacylglycerols were lower in apo-B48-only mice and higher in the apo-B100-only mice. In the setting of apo-E-deficiency, the apo-B100-only mutation lowered cholesterol levels, consistent with the fact that apo-B100-lipoproteins can be cleared from the plasma via the LDL receptor, whereas apo-B48-lipoproteins lacking apo-E cannot. The apo-B48-only and apo-B100-only mice should prove to be valuable models for experiments designed to understand the purpose for the two forms of apo-B in mammalian metabolism.

DNA ligase I mediates essential functions in mammalian cells

DNA replication, repair, and recombination are essential processes in mammalian cells. Hence, the application of gene targeting to the study of these DNA metabolic pathways requires the creation of nonnull mutations. We have developed a method for introducing partially defective mutants in murine embryonic stem cells that circumvents the problem of cellular lethality of targeted mutations at essential loci. Using this approach, we have determined that mammalian DNA ligase I is essential for cell viability. Thus, DNA ligases II and III are not redundant with DNA ligase I for the function(s) associated with cell proliferation. Partial complementation of the lethal DNA ligase I null mutation allowed the creation of deficient embryonic stem cell lines. We found that a wild-type DNA ligase I cDNA, as well as a variant DNA ligase I cDNA, was able to rescue the lethality of the homozygous null mutation, whereas an N-terminal deletion mutant consisting of the minimal DNA ligase I catalytic domain was not. This observation demonstrates that sequences outside the DNA ligase I catalytic domain are essential for DNA ligase I function in vivo.

High-frequency gene conversion between repeated C mu sequences integrated at the chromosomal immunoglobulin mu locus in mouse hybridoma cells

The occurrence of mitotic recombination between repeated immunoglobulin mu gene constant (C mu) region sequences stably integrated at the haploid chromosomal immunoglobulin mu locus in murine hybridoma cells was investigated. Recombination events are detected as changes in hapten-specific immunoglobulin M production. Recombination occurs with high frequency (0.5 to 0.8%) by a mechanism consistent with gene conversion. A double-strand break repair-like mechanism is suggested by the finding that repair of a 2-bp deletion mutation and a 2-bp insertion mutation occurs with parity in a donor-directed manner. The results also suggest that the gene conversion process is directional in that the 5' C mu region sequence is preferentially converted.

Double replacement: strategy for efficient introduction of subtle mutations into the murine Col1a-1 gene by homologous recombination in embryonic stem cells

A subtle mutation that rendered type I collagen resistant to mammalian collagenase has been introduced into the murine Col1a-1 (recently redesignated Cola-1) gene by homologous recombination in embryonic stem (ES) cells. Initially, a "hit and run" procedure was used. Since two steps were required for introducing each mutation and more than one mutation was to be introduced in the same genomic region independently, we have developed a streamlined procedure that involves two sequential replacement-type homologous recombination events. In the first step, an internal deletion was introduced into the Col1a-1 locus along with the positive and negative selectable markers, neo and tk, to mark the region of interest. G418-resistant homologous recombinants were isolated and used in the second step in which the deleted Col1a-1 allele was replaced with a construct containing the desired mutation. Homologous recombinants containing the mutation were identified among the Tk- ES clones after selection with FIAU [1-(2-deoxy-2-fluoro-beta-D-arabinofuranosyl)-5-iodouracil (called fialuridine)]. Approximately 10% of such clones contained the desired mutation. The double replacement procedure greatly reduces the time and amount of work required to introduce mutations independently into the same or closely linked regions. Once the homologous recombinants derived from the first step are established, the introduction of other mutations into the deleted region becomes a one-step procedure. For X number of introduced mutations, 2X selections are required with the "hit and run" approach, but only X + 1 are required with the double-replacement method. This innovative procedure could be very useful in studies of gene structure and function as well as gene expression and regulation.

Use of double-replacement gene targeting to replace the murine alpha-lactalbumin gene with its human counterpart in embryonic stem cells and mice

The mouse alpha-lactalbumin gene has been replaced with the human gene by two consecutive rounds of gene targeting in hypoxanthine phosphoribosyltransferase (HPRT)-deficient feeder-independent murine embryonic stem (ES) cells. One mouse alpha-lactalbumin allele was first replaced by an HPRT minigene which was in turn replaced by human alpha-lactalbumin. The end result is a clean exchange of defined DNA fragments with no other DNA remaining at the target locus. Targeted ES cells at each stage remained capable of contributing efficiently to the germ line of chimeric animals. Double replacement using HPRT-deficient ES cells and the HPRT selection system is therefore a powerful and flexible method of targeting specific alterations to animal genes. A typical strategy for future use would be to generate a null mutation which could then be used to produce multiple second-step alterations at the same locus.

Use of double-replacement gene targeting to replace the murine alpha-lactalbumin gene with its human counterpart in embryonic stem cells and mice

The mouse alpha-lactalbumin gene has been replaced with the human gene by two consecutive rounds of gene targeting in hypoxanthine phosphoribosyltransferase (HPRT)-deficient feeder-independent murine embryonic stem (ES) cells. One mouse alpha-lactalbumin allele was first replaced by an HPRT minigene which was in turn replaced by human alpha-lactalbumin. The end result is a clean exchange of defined DNA fragments with no other DNA remaining at the target locus. Targeted ES cells at each stage remained capable of contributing efficiently to the germ line of chimeric animals. Double replacement using HPRT-deficient ES cells and the HPRT selection system is therefore a powerful and flexible method of targeting specific alterations to animal genes. A typical strategy for future use would be to generate a null mutation which could then be used to produce multiple second-step alterations at the same locus.

The frequency of gene targeting in yeast depends on the number of target copies

We have compared the efficiency of transformation by linear DNA fragments in yeast strains carrying different numbers of homologous targets for recombination. In strains carrying dispersed copies of a target and in strains carrying tandem arrays, the frequency of transformation is proportional to the number of targets. This result is in contrast to previous studies of transformation in mammalian cells, where targeted integration was insensitive to the number of targets. We conclude that, in yeast, the search for a homologous partner is a rate-limiting step in the successful recombination of linearized DNA fragments. Furthermore, the fact that we obtain the same results with both dispersed and clustered targets argues against models of homology searching in which DNA becomes nonspecifically associated with a chromosome and then slides along the DNA until homology is encountered.

An "in-out" strategy using gene targeting and FLP recombinase for the functional dissection of complex DNA regulatory elements: analysis of the beta-globin locus control region

The human beta-globin locus control region (LCR) is a complex DNA regulatory element that controls the expression of the cis-linked beta-like globin genes located in the 55 kilobases 3' of the LCR. We have initiated the functional analysis of the LCR by homologous recombination in murine erythroleukemia cell somatic hybrids that carry a single copy of human chromosome 11 on which the beta-globin locus is situated. High-level expression of the human beta-globin gene normally occurs when these hybrid cells are induced to differentiate. We have reported that the insertion of an expressed selectable marker gene (driven by the Friend virus enhancer/promoter) into the LCR disrupts the LCR-mediated regulation of globin transcription. In these cells, beta-globin is no longer expressed when the cells differentiate; instead, expression of the selectable marker gene increases significantly after differentiation. Since present techniques for homologous recombination require the insertion of a selectable marker, further progress in using homologous recombination to analyze the LCR depends on deletion of the selectable marker and demonstration that the locus functions normally after the insertion, expression, and deletion of the selectable marker. Here we show that after precise deletion of the selectable marker by using the FLP recombinase/FRT (FLP recombinase target) system, the locus functions as it did before the homologous recombination event. These studies demonstrate the feasibility of using homologous recombination to analyze the LCR in particular, and other complex cis-regulatory DNA elements in general, in their normal chromosomal context.

Site-directed point mutations in embryonic stem cells: a gene-targeting tag-and-exchange strategy

Sequential gene targeting was used to introduce point mutations into one alpha 2 isoform Na,K-ATPase homolog in mouse embryonic stem (ES) cells. In the first round of targeted replacement, the gene was tagged with selectable markers by insertion of a Neor/HSV-tk gene cassette, and this event was selected for by gain of neomycin (G418) resistance. In the second targeted replacement event, the tagged genomic sequence was exchanged with a vector consisting of homologous genomic sequences carrying five site-directed nucleotide substitutions. Embryonic stem cell clones modified by exchange with the mutation vector were selected for loss of the HSV-tk gene by resistance to ganciclovir. Candidate clones were further screened and identified by polymerase chain reaction and Southern blot analysis. By this strategy, the endogenous alpha 2 isoform Na,K-ATPase gene was altered to encode two other amino acids so that the enzyme is resistant to inhibition by cardiac glycosides while maintaining its transmembrane ion-pumping function. Since the initial tagging event and the subsequent mutation-exchange event are independent of one another, a tagged cell line can be used to generate a variety of mutant lines by exchange with various mutation vectors at the tagged locus. This method should be useful for testing specific mutations introduced into the genomes of tissue culture cells and animals and for developing animal models encompassing the mutational variability of known genetic disorders.

Site-directed point mutations in embryonic stem cells: a gene-targeting tag-and-exchange strategy

Sequential gene targeting was used to introduce point mutations into one alpha 2 isoform Na,K-ATPase homolog in mouse embryonic stem (ES) cells. In the first round of targeted replacement, the gene was tagged with selectable markers by insertion of a Neor/HSV-tk gene cassette, and this event was selected for by gain of neomycin (G418) resistance. In the second targeted replacement event, the tagged genomic sequence was exchanged with a vector consisting of homologous genomic sequences carrying five site-directed nucleotide substitutions. Embryonic stem cell clones modified by exchange with the mutation vector were selected for loss of the HSV-tk gene by resistance to ganciclovir. Candidate clones were further screened and identified by polymerase chain reaction and Southern blot analysis. By this strategy, the endogenous alpha 2 isoform Na,K-ATPase gene was altered to encode two other amino acids so that the enzyme is resistant to inhibition by cardiac glycosides while maintaining its transmembrane ion-pumping function. Since the initial tagging event and the subsequent mutation-exchange event are independent of one another, a tagged cell line can be used to generate a variety of mutant lines by exchange with various mutation vectors at the tagged locus. This method should be useful for testing specific mutations introduced into the genomes of tissue culture cells and animals and for developing animal models encompassing the mutational variability of known genetic disorders.

A series of vectors that simplify mammalian gene targeting

In order to facilitate the procedure of mammalian gene targeting, we have produced and functionally tested a series of generic vectors. Homologous recombination has been achieved with each vector. The vectors are designed for both replacement and insertional recombination, are suitable for 'hit and run' strategies and contain all necessary genetic elements for both positive-negative and promoterless/gene fusion enrichment of homologous integrations. Multiple unique restriction sites are included to simplify the incorporation of genomic targeting sequences.

Introduction of a point mutation into the mouse genome by homologous recombination in embryonic stem cells using a replacement type vector with a selectable marker

The introduction of small mutations instead of null alleles into the mouse genome has broad applications to the study of protein structure-function relationships and the creation of animal models of human genetic diseases. To test a simple mutational strategy we designed a targeting vector for the mouse proopiomelanocortin (POMC) gene containing a single nucleotide insertion that converts the initial tyrosine codon of beta-endorphin 1-31 to a premature translational termination codon and introduces a unique Hpal endonuclease restriction site. The targeting vector also contains a neo cassette immediately 3' to the last POMC exon and a herpes simplex virus thymidine kinase cassette to allow positive and negative selection. Homologous recombination occurred at a frequency of 1/30 clones of electroporated embryonic stem cells selected in G418 and gancyclovir. 10/11 clones identified initially by a polymerase chain reaction (PCR) strategy had the predicted structure without evidence of concatemer formation by Southern blot analysis. We used a combination of Hpa I digestion of PCR amplified fragments and direct nucleotide sequencing to further confirm that the point mutation was retained in 9/10 clones. The POMC gene was transcriptionally silent in embryonic stem cells and the targeted allele was not activated by the downstream phosphoglycerate kinase-1 promoter that transcribed the neo gene. Under the electroporation conditions used, we have demonstrated that a point mutation can be introduced with high efficiency and precision into the POMC gene using a replacement type vector containing a retained selectable marker without affecting expression of the allele in the embryonic stem cells. A similar strategy may be useful for a wide range of genes.

Retrovirus-mediated insertion of expressed and non-expressed genes at identical chromosomal locations

During retrovirus replication, a cellularly derived tRNA is annealed to the viral RNA at the primer binding site (PBS) to prime reverse transcription, and both the tRNA and the PBS become copied and matched together on complementary proviral DNA strands prior to integration. Using a viral PBS single base pair mutant which affects provirus expression in undifferentiated cells, we show that reversion to wild type (wt) occurs at a frequency of approximately 50%. Daughter cell lines containing wt or mutant proviruses at identical chromosomal sites have been isolated, supporting a model where an integrated PBS-mismatched provirus was copied before mismatch correction could occur. Virus expression in daughter cells containing the mutant provirus was 100-fold higher than in cells bearing the wt counterpart. Additionally, proviral 5' DNA and cellular 5' flanking DNA became methylated in daughter cells containing wt but not mutant integrants. These results strongly support the current model of retrovirus reverse transcription, and indicate that the wt PBS region contains an element which suppresses virus expression and directs the methylation of viral and neighboring cellular DNA.

Location of crossovers during gene targeting with insertion and replacement vectors

Gene targeting was used to introduce nonselectable genetic changes into chromosomal loci in mouse embryo-derived stem cells. The nonselectable markers were linked to a selectable marker in both insertion- and replacement-type vectors, and the transfer of the two elements to the Hprt locus was assayed. When insertion vectors were used as substrates, the frequency of transfer was highly dependent upon the distance between the nonselectable marker and the double-strand break in the vector. A marker located close to the vector ends was frequently lost, suggesting that a double-strand gap repair activity is involved in vector integration. When replacement vectors were used, cotransfer of a selectable marker and a nonselectable marker 3 kb apart was over 50%, suggesting that recombination between vector and target often occurs near the ends of the vector. To illustrate the use of replacement vectors to transfer specific mutations to the genome, we describe targeting of the delta F508 mutation to the CFTR gene in mouse embryo-derived stem cells.

The effects of terminal heterologies on gene targeting by insertion vectors in embryonic stem cells

We have examined the effects of placing nonhomologous DNA on the ends of an insertion-type gene targeting vector. The presence of terminal heterologies was found to be compatible with insertion targeting, and the terminal heterologies were efficiently removed. Terminal heterologies reduced the frequency of gene targeting to variable extents. The degree of inhibition of targeting was dependent on the length and the position of the heterology: 2.1kb heterologous sequences were more inhibitory than shorter regions of heterology, and heterology placed on the end of the long (4.8kb) arm of homology was more inhibitory than heterology positioned on the end of the short (0.8kb) arm. When heterology was placed on both arms of the targeting vector the targeting efficiencies were similar to or higher than when heterology was present on the long arm only. These results suggest that terminal sequences are removed simultaneously from both ends of targeting vectors. The removal of terminal sequences probably occurs by exonucleolytic degradation of both strands at each end, and removal of at least one of the strands is intimately coupled with the process of homologous recombination. These findings have implications for the design of gene targeting vectors.

Location of crossovers during gene targeting with insertion and replacement vectors

Gene targeting was used to introduce nonselectable genetic changes into chromosomal loci in mouse embryo-derived stem cells. The nonselectable markers were linked to a selectable marker in both insertion- and replacement-type vectors, and the transfer of the two elements to the Hprt locus was assayed. When insertion vectors were used as substrates, the frequency of transfer was highly dependent upon the distance between the nonselectable marker and the double-strand break in the vector. A marker located close to the vector ends was frequently lost, suggesting that a double-strand gap repair activity is involved in vector integration. When replacement vectors were used, cotransfer of a selectable marker and a nonselectable marker 3 kb apart was over 50%, suggesting that recombination between vector and target often occurs near the ends of the vector. To illustrate the use of replacement vectors to transfer specific mutations to the genome, we describe targeting of the delta F508 mutation to the CFTR gene in mouse embryo-derived stem cells.

Modification of the apolipoprotein B gene in HepG2 cells by gene targeting

The HepG2 cell line has been used extensively to study the synthesis and secretion of apolipoprotein (apo) B. In this study, we tested whether gene-targeting techniques can be used to inactivate one of the apo B alleles in HepG2 cells by homologous recombination using a transfected gene-targeting vector. Our vector contained exons 1-7 of the apo B gene, in which exon 2 was interrupted by a promoterless neomycin resistance (neo(r)) gene. The recombination of this vector with the cognate gene would inactivate an apo B allele and enable the apo B promoter to activate the transcription of the neo(r) gene. To detect the rare homologous recombinant clone, we developed a novel solid phase RIA that uses the apo B-specific monoclonal antibody MB19 to analyze the apo B secreted by G418-resistant (G418r) clones. Antibody MB19 detects a two-allele genetic polymorphism in apo B by binding to the apo B allotypes MB19(1) and MB19(2) with high and low affinity, respectively. HepG2 cells normally secrete both the apo B MB19 allotypes. Using the MB19 immunoassay, we identified a G418r HepG2 clone that had lost the ability to secrete the MB19(1) allotype. The inactivation of an apo B allele of this clone was confirmed by the polymerase chain reaction amplification of an 865-bp fragment unique to the targeted apo B allele and by Southern blotting of genomic DNA. This study demonstrates that gene-targeting techniques can be used to modify the apo B gene in HepG2 cells and demonstrates the usefulness of a novel solid phase RIA system for detecting apo B gene targeting events in this cell line.

Disruption of the cystic fibrosis transmembrane conductance regulator gene in embryonic stem cells by gene targeting

We have successfully disrupted the cftr (cystic fibrosis transmembrane conductance regulator) gene at its endogenous locus in embryonic stem cells by gene targeting. We are using a double replacement strategy to introduce subtle mutations into exon 10. We report here the first step of creating a null mutation by insertion of a functional hprt (hypoxanthine phosphoribosyl transferase) mini-gene into exon 10 of the cftr gene. Targeted embryonic stem cell clones were identified by PCR screening and confirmed by Southern blot analysis. One of the cftr targeted clones has been injected into recipient blastocysts and shown to contribute to chimaeras. The targeted clones will now be used as the starting point for a second gene targeting step to remove the hprt gene in exon 10 with the concomitant introduction of the delta F508 mutation or other mutations.

Investigation of coelectroporation as a method for introducing small mutations into embryonic stem cells

We have investigated coelectroporation as a method for introducing minor genetic changes into specific genes in embryonic stem cells. A selectable marker (neo) and a targeting replacement vector designed to insert a 4-bp insertion into exon 3 of the mouse hypoxanthine-guanine phosphoribosyltransferase (HPRT) gene were coelectroporated into embryonic stem cells and selected in G418 and 6-thioguanine (6-TG). HPRT-negative clones were obtained at a frequency of approximately 1 per 520 G418r clones. Southern analysis and the polymerase chain reaction were used to demonstrate that 3 of 36 of the 6-TG-resistant clones had the desired 4-bp insertion without any other disruption of the HPRT locus. Initial studies indicated that the other 33 6-TG-resistant clones probably resulted from the targeted integration of a concatemer containing both the targeting construct and the selectable neo gene.

Investigation of coelectroporation as a method for introducing small mutations into embryonic stem cells

We have investigated coelectroporation as a method for introducing minor genetic changes into specific genes in embryonic stem cells. A selectable marker (neo) and a targeting replacement vector designed to insert a 4-bp insertion into exon 3 of the mouse hypoxanthine-guanine phosphoribosyltransferase (HPRT) gene were coelectroporated into embryonic stem cells and selected in G418 and 6-thioguanine (6-TG). HPRT-negative clones were obtained at a frequency of approximately 1 per 520 G418r clones. Southern analysis and the polymerase chain reaction were used to demonstrate that 3 of 36 of the 6-TG-resistant clones had the desired 4-bp insertion without any other disruption of the HPRT locus. Initial studies indicated that the other 33 6-TG-resistant clones probably resulted from the targeted integration of a concatemer containing both the targeting construct and the selectable neo gene.

Polarized and functional epithelia can form after the targeted inactivation of both mouse keratin 8 alleles

We have tested the requirement of keratin intermediate filaments for the formation and function of a simple epithelium. We disrupted both alleles of the mouse keratin 8 (mK8) gene in embryonic stem cells, and subsequently analyzed the phenotype in developing embryoid bodies in suspension culture. After the inactivation of the mouse keratin 8 (mK8) gene by a targeted insertion, mK8 protein synthesis was undetectable. In the absence of mK8 its complementary partners mK18 and mK19 were unable to form filaments within differentiated cells. Surprisingly, these ES cells differentiate to both simple and cystic embryoid bodies with apparently normal epithelia. Ultrastructural analysis shows an apparently normal epithelium with microvilli on the apical membrane, tight junctions and desmosomes on the lateral membrane, and an underlying basal membrane. No significant differences in the synthesis or secretion of alpha 1-fetoprotein and laminin were observed between the mK8- or wild-type embryoid bodies. Our data show that mK8 is not required for simple epithelium formation of extraembryonic endoderm.

A dominant positive and negative selectable gene for use in mammalian cells

We have constructed three different fusion genes containing the herpes simplex virus thymidine kinase (HSV tk) and the bacterial neomycin phosphotransferase (neo) genes. All three fusion genes utilize the HSV tk promoter but differ at the junction of their components. We have determined if the fusion genes are bifunctional by introducing them into mammalian cells and testing for function of the individual components. One of the fusion genes, TNFUS 69, produced a bicistronic message and a fusion protein that has TK and NEO protein functions. This and other fusion genes of a similar nature could serve as dominant positive and negative selectable markers in mammalian cells.

Toward an animal model of cystic fibrosis: targeted interruption of exon 10 of the cystic fibrosis transmembrane regulator gene in embryonic stem cells

A gene-targeting construct was made containing 7.8 kilobases of DNA spanning exon 10 of the mouse cystic fibrosis transmembrane regulator (CFTR) gene in which part of the exon has been replaced by two neomycin-resistance (Neo) genes driven by different promoters. (This replacement introduces a chain-termination codon at amino acid position 489 in the CFTR sequence). A herpes simplex thymidine kinase gene was on each end of the construct, which was electroporated into embryonic stem (ES) cells. Colonies resistant to G418, or to G418 plus ganciclovir, were selected and screened by Southern blotting or by PCR amplification. Five pools of G418-resistant cells gave PCR products diagnostic of targeting. Four independent clones of ES cells with a disrupted CFTR gene have been isolated from these pools. The frequency of targeting was 1/2500 G418-resistant colonies. This low frequency is not the consequence of marginal expression of the Neo genes in the targeted cells. The CFTR targeting events were clustered among our experiments in a manner suggesting that some unidentified factor(s), possibly passage number, influences the recovery of CFTR-targeted cells.

Double-strand gap repair in a mammalian gene targeting reaction

To better understand the mechanism of homologous recombination in mammalian cells that facilitates gene targeting, we have analyzed the recombination reaction that inserts a plasmid into a homologous chromosomal locus in mouse embryonic stem cells. A partially deleted HPRT gene was targeted with various plasmids capable of correcting the mutation at this locus, and HPRT+ recombinants were directly selected in HAT medium. The structures of the recombinant loci were then determined by genomic Southern blot hybridizations. We demonstrate that plasmid gaps of 200, 600, and 2,500 bp are efficiently repaired during the integrative recombination reaction. Targeting plasmids that carry a double-strand break or gap in the region of DNA homologous to the target locus produce 33- to 140-fold more hypoxanthine-aminopterin-thymidine-resistant recombinants than did these same plasmids introduced in their uncut (supercoiled) forms. Our data suggest that double-strand gaps and breaks may be enlarged prior to the repair reaction since sequence heterologies carried by the incoming plasmids located close to them are often lost. These results extend the known similarities between mammalian and yeast recombination mechanisms and suggest several features of the insertional (O-type) gene targeting reaction that should be considered when one is designing mammalian gene targeting experiments.

Fidelity of targeted recombination in human fibroblasts and murine embryonic stem cells

Targeted recombination in murine embryonic stem cells promises to be a powerful tool for introducing specific mutations into target genes to study development in mice and to create animal models of human disease. Gene targeting also holds potential for correcting genetic defects as an approach to human gene therapy. To precisely modify target genes, homologous recombination must proceed with high fidelity. However, several results have suggested that targeted recombination may be highly mutagenic. To test the accuracy of gene targeting we analyzed 44 independent targeted recombinants at the hypoxanthine phosphoribosyltransferase (HPRT) locus in a human fibroblast cell line and in mouse embryonic stem cells. We surveyed 80 kilobases around the sites of recombination by using chemical cleavage of mismatches. Only two mutations were found: a T----G transversion and a thymidine deletion. Thus, gene targeting in mammalian cells can be extremely accurate. These results demonstrate the feasibility of generating precise modifications of mammalian genomes by gene targeting.

Double-strand gap repair in a mammalian gene targeting reaction

To better understand the mechanism of homologous recombination in mammalian cells that facilitates gene targeting, we have analyzed the recombination reaction that inserts a plasmid into a homologous chromosomal locus in mouse embryonic stem cells. A partially deleted HPRT gene was targeted with various plasmids capable of correcting the mutation at this locus, and HPRT+ recombinants were directly selected in HAT medium. The structures of the recombinant loci were then determined by genomic Southern blot hybridizations. We demonstrate that plasmid gaps of 200, 600, and 2,500 bp are efficiently repaired during the integrative recombination reaction. Targeting plasmids that carry a double-strand break or gap in the region of DNA homologous to the target locus produce 33- to 140-fold more hypoxanthine-aminopterin-thymidine-resistant recombinants than did these same plasmids introduced in their uncut (supercoiled) forms. Our data suggest that double-strand gaps and breaks may be enlarged prior to the repair reaction since sequence heterologies carried by the incoming plasmids located close to them are often lost. These results extend the known similarities between mammalian and yeast recombination mechanisms and suggest several features of the insertional (O-type) gene targeting reaction that should be considered when one is designing mammalian gene targeting experiments.

Prospects for homologous recombination in human gene therapy

The ideal approach to gene therapy of hereditary diseases or gene correction therapy is considered. The advantages, disadvantages and limits of gene targeting by homologous recombination are discussed with regard to its possible application in gene correction therapy and in comparison with retroviral-mediated gene complementation therapy.