Targeted modification of a human beta-globin locus BAC clone using GET Recombination and an I-Scei counterselection cassette

There is a need for better approaches to allow precise engineering of large genomic BAC DNA fragments, to facilitate the use of intact genomic loci for therapeutic and biotechnology applications. We report an efficient method to insert any modification in any genomic locus, using a human beta-globin locus BAC clone as a model system. The modifications can range from single base changes to large insertions or deletions and leave no operational sequences. A counterselection cassette, consisting of an inducible I-SceI gene, its recognition site, and an antibiotic resistance gene, is inserted into the targeted region using GET Recombination. A PCR fragment carrying the modification but no selectable marker replaces the counterselection cassette in a second round of GET Recombination. The unique I-SceI site in the counterselection cassette is cut by I-SceI endonuclease, strongly selecting against nonrecombinant clones and yielding up to 30% correct recombinants.     

Engineering EGFP reporter constructs into a 200 kb human beta-globin BAC clone using GET Recombination

GET Recombination, a simple inducible homologous recombination system for Escherichia coli, was used to target insertion of an EGFP cassette between the start and termination codons of the beta-globin gene in a 200 kb BAC clone. The high degree of homology between the promoter regions of the beta- and delta-globin genes also allowed the simultaneous generation of a delta-globin reporter construct with the deletion of 8.8 kb of intervening sequences. Both constructs expressed EGFP after transient transfection of MEL cells. Similarly, targeting of the EGFP cassette between the promoter regions of the gamma-globin genes and the termination codon of the beta-globin gene enabled the generation of reporter constructs for both (A)gamma- and (G)gamma-globin genes, involving specific deletions of 24 and 29 kb of genomic sequence, respectively. Finally the EGFP cassette was also inserted between the epsilon- and beta-globin genes, with the simultaneous deletion of 44 kb of intervening sequence. The modified constructs were generated at high efficiency, illustrating the usefulness of GET Recombination to generate large deletions of specific sequences in BACs for functional studies. The establishment of stable erythropoietic cell lines with these globin constructs will facilitate the search for therapeutic agents that modify the expression of the individual globin genes in a physiologically relevant manner.     

Insertion of modifications in the beta-globin locus using GET recombination with single-stranded oligonucleotides and denatured PCR fragments

We describe the use of the GET recombination system with oligonucleotides or single-stranded polymerase chain reaction (PCR) fragments to insert modifications in the human beta-globin locus without counterselection. The method involves recombination between oligonucleotides or denatured PCR fragments and homologous sequences in the beta-globin gene in a clone of 205-kb bacterial artificial chromosome (BAC), based on the inducible expression of the recE, recT, and gam genes. In this method, oligonucleotides or denatured PCR fragments are electroporated directly into cells carrying both the globin BAC and the pGETrec plasmid, after induction of the GET recombination system. Recombinant BAC clones are identified by PCR, using allele-specific amplification for the mutated sequences. We have used this approach to insert a unique restriction site as well as a common thalassemia mutation (stop codon 39, C-->T) into the human beta-globin locus. We have observed the frequency of recombinant clones to be as high as 1 in 100-200 clones. Therefore, this approach provides a simple and efficient method for introducing point mutations and other fine modifications into BACs, and should greatly facilitate the use of BACs for functional studies and therapeutic applications.     

Clustering of null mutations in the EcoRI endonuclease

EcoRI endonuclease mutants were isolated in a methylase-deficient background following in vitro hydroxylamine mutagenesis of plasmid pKG2 (Kuhn et al.: Gene 44:253-263, 1986). Mutants which survived high-level endonuclease expression (IPTG induction) were termed null mutants. Sixty-two of 121 null mutants tested by Western blot contained normal levels of endonuclease cross-reacting protein. The complete endonuclease gene was sequenced for 27 null mutants. This group was found to consist of 20 single base-change missense mutations, 6 double mutations, and 1 triple mutation. Ten of the 20 single mutations were clustered between residues 139 and 144. When examined with respect to the structure of the EcoRI-DNA complex (McClarin et al.: Science 234:1526-1541, 1986), these alterations were found to fall predominantly into two classes: substitutions at the protein-DNA interface or substitutions at the protein-protein (dimer) interface. Protein from several of the mutants was purified and sized by using HPLC. Wild-type EcoRI endonuclease and protein from three of the DNA interface mutations (Ala139----Thr, Gly140----Ser, Arg203----Gln) appeared to be dimeric, while protein from subunit interface mutations (Glu144----Lys, Glu152----Lys, Gly210----Arg) migrated as monomers.     

Engineering EGFP reporter constructs into a 200 kb human β-globin BAC clone using GET Recombination

GET Recombination, a simple inducible homologous recombination system for Escherichia coli, was used to target insertion of an EGFP cassette between the start and termination codons of the β-globin gene in a 200 kb BAC clone. The high degree of homology between the promoter regions of the β- and δ-globin genes also allowed the simultaneous generation of a δ-globin reporter construct with the deletion of 8.8 kb of intervening sequences. Both constructs expressed EGFP after transient transfection of MEL cells. Similarly, targeting of the EGFP cassette between the promoter regions of the γ-globin genes and the termination codon of the β-globin gene enabled the generation of reporter constructs for both ^Aγ- and ^Gγ-globin genes, involving specific deletions of 24 and 29 kb of genomic sequence, respectively. Finally the EGFP cassette was also inserted between the - and β-globin genes, with the simultaneous deletion of 44 kb of intervening sequence. The modified constructs were generated at high efficiency, illustrating the usefulness of GET Recombination to generate large deletions of specific sequences in BACs for functional studies. The establishment of stable erythropoietic cell lines with these globin constructs will facilitate the search for therapeutic agents that modify the expression of the individual globin genes in a physiologically relevant manner.     

Insertion of a rabbit beta-globin gene sequence into an E. coli plasmid.

Double stranded DNA has been synthesized in vitro from rabbit globin messenger RNA and elongated with homopolymeric dG tails. An E. coli plasmid was cleaved by EcoRI. The cohesive ends were repaired and dC tails added, to permit reconstitution of the EcoRI sites upon annealing with the dG elongated globin DNA. Transformation of E. coli with the globin-plasmid DNA hybrid has yielded a clone which harbours a recombinant plasmid (pCR1-betaG1), as demonstrated by hybridization experiments with radioactive globin cDNA. The sequence carried by the recombinant plasmid corresponds to part of the gene sequence coding for the beta chain of rabbit globin. Circular DNA of the purified recombinant plasmid exhibits sensitivity to EcoRI.     

Solubility of the EcoRI restriction endonuclease and its purification from an over-producing strain

The solubility of the EcoRI restriction endonuclease was measured in solutions of varying NaCl concentrations, at different temperatures and in the presence of DNA. The precipitation of the protein was enhanced by low NaCl concentrations, by elevated temperatures, and by the addition of DNA. These observations are discussed in relationship to the interaction of this protein with DNA. The purification of the EcoRI restriction enzyme from a strain of Escherichia coli that over-produces this enzyme was hampered by the insolubility of the protein, and hence the purification procedure was modified to optimize the recovery of active enzyme.     

Haplotype analysis of the human beta-globin gene complex using multiple locus specific oligonucleotide probes

Three oligonucleotide probes complementary to specific DNA sequences of the six human globin genes (epsilon, G gamma, A gamma, psi beta, delta, beta) were synthesized. The oligonucleotides were used either singly or in combination as hybridization probes to determine the haplotype of the human beta-globin gene cluster employing the four conventionally used restriction endonucleases HincII, HindIII, AvaII, and BamHI, in addition to HpaI. Polymorphism in the epsilon- and psi beta-genes (HincII) can be simultaneously determined with a single probe mixture. One of the probes complementary to both the psi beta- and gamma-genes is useful for determining both HindIII and HincII polymorphisms. The advantages of these probes relative to conventional cDNA probes are discussed.     

Purification and characterization of the restriction endonuclease RsrI, an isoschizomer of EcoRI

Rhodobacter sphaeroides strain 630 produces restriction enzyme RsrI which is an isoschizomer of EcoRI. We have purified this enzyme and initiated a comparison with the EcoRI endonuclease. The properties of RsrI are consistent with a reaction mechanism similar to that of EcoRI: the position of cleavage within the -GAATTC-site is identical, the MgCl2 optimum for the cleavage is identical, and the pH profile is similar. Methylation of the substrate sequence by the EcoRI methylase protects the site from cleavage by the RsrI endonuclease. RsrI cross-reacts strongly with anti-EcoRI serum indicating three-dimensional structural similarities. We have determined the sequence of 34 N terminal amino acids for RsrI and this sequence possesses significant similarity to the EcoRI N terminus.     

Multiple sites of replication initiation in the human beta-globin gene locus

The cell cycle-dependent, ordered assembly of protein prereplicative complexes suggests that eukaryotic replication origins determine when genomic replication initiates. By comparison, the factors that determine where replication initiates relative to the sites of prereplicative complex formation are not known. In the human globin gene locus previous work showed that replication initiates at a single site 5′ to the β-globin gene when protein synthesis is inhibited by emetine. The present study has examined the pattern of initiation around the genetically defined β-globin replicator in logarithmically growing HeLa cells, using two PCR-based nascent strand assays. In contrast to the pattern of initiation detected in emetine-treated cells, analysis of the short nascent strands at five positions spanning a 40 kb globin gene region shows that replication initiates at more than one site in non-drug-treated cells. Quantitation of nascent DNA chains confirmed that replication begins at several locations in this domain, including one near the initiation region (IR) identified in emetine-treated cells. However, the abundance of short nascent strands at another initiation site ~20 kb upstream is ~4-fold as great as that at the IR. The latter site abuts an early S phase replicating fragment previously defined at low resolution in logarithmically dividing cells.     

The specific binding, bending, and unwinding of DNA by RsrI endonuclease, an isoschizomer of EcoRI endonuclease

To determine whether RsrI endonuclease recognizes and cleaves the sequence GAATTC in duplex DNA similarly to its isoschizomer EcoRI we initiated a functional comparison of the two enzymes. Equilibrium binding experiments showed that at 20 degrees C RsrI endonuclease binds to specific and nonspecific sequences in DNA with affinities similar to those of EcoRI. At 0 degrees C the affinity of RsrI for its specific recognition sequence is reduced 7-fold whereas the affinity for noncanonical sequences remains relatively unchanged. Unlike EcoRI, incubation of RsrI endonuclease with N-ethylmaleimide inactivates the enzyme; however, preincubation with DNA prevents the inactivation. The N-ethylmaleimide-treated enzyme fails to bind DNA as assayed by gel mobility shift assays. Comparison of the deduced amino acid sequences of RsrI and EcoRI endonucleases suggests that modification of Cys245 is responsible for the inactivation. Fe(II). EDTA and methidiumpropyl-EDTA.Fe(II) footprinting results indicate that RsrI, like EcoRI, protects 12 base pairs from cleavage when bound to its specific recognition sequence in the absence of Mg2+. RsrI bends DNA by approximately 50 degrees, as determined by measuring the relative electrophoretic mobilities of specific RsrI-DNA complexes with the binding site in the center or near the end of the DNA fragment. This value is similar to that reported for EcoRI. RsrI also unwinds the DNA helix by 25 degrees +/- 5 degrees, a value close to that reported for EcoRI endonuclease. Collectively, these results indicate that the overall structural changes induced in the DNA by the binding of RsrI and EcoRI endonucleases to DNA in the absence of Mg2+ are similar. In the accompanying paper (Aiken, C. R., McLaughlin, L. W., and Gumport, R. I. (1991) J. Biol. Chem. 266, 19070-19078) we present results of studies of RsrI endonuclease using oligonucleotide substrates containing base analogues which suggest differences in the ways the two enzymes cleave DNA.     

Probing the role of glutamic acid 144 in the EcoRI endonuclease using aspartic acid and glutamine replacements

The x-ray structure of the EcoRI endonuclease-DNA complex (3) suggests that hydrogen bonds between amino acids, glutamic acid 144, arginine 145, and arginine 200, and major groove base moieties are the molecular determinants of specificity. We have investigated residue 144 using aspartate and glutamine substitutions introduced by site-directed mutagenesis. Substitution with glutamine results in a null phenotype (at least a 2000-fold reduction in activity). On the other hand, the aspartic acid mutant (ED144) retained in vivo activity. Substrate binding and catalytic studies were done with purified ED144 enzyme. The affinity of the ED144 enzyme for the canonical sequence 5'-GAATTC-3' is about 340-fold less than the wild-type (WT) enzyme, while its affinity for nonspecific DNA is about 50 times greater. The ED144 enzyme cleaves one strand in the EcoRI site in plasmid pBR322 with a kcat/Km similar to WT. In contrast to the WT enzyme, the ED144 enzyme dissociates after the first strand cleavage. Partitioning between cleavage and dissociation at the first and second cleavage steps for the ED144 enzyme is extremely salt-sensitive. The altered partitioning results largely from a destabilization of the enzyme-DNA complex, particularly the enzyme-nicked DNA complex, with only small changes in the respective cleavage rates. The hydrogen bonds of Glu-144 are critical, they appear to act cooperatively with other specificity contacts to stabilize the enzyme-DNA complex.     

Stimulation of homology-directed gene targeting at an endogenous human locus by a nicking endonuclease

Homologous recombination (HR) is a highly accurate mechanism of DNA repair that can be exploited for homology-directed gene targeting. Since in most cell types HR occurs very infrequently (∼10⁻⁶ to 10⁻⁸), its practical application has been largely restricted to specific experimental systems that allow selection of the few cells that become genetically modified. HR-mediated gene targeting has nonetheless revolutionized genetics by greatly facilitating the analysis of mammalian gene function. Recent studies showed that generation of double-strand DNA breaks at specific loci by designed endonucleases greatly increases the rate of homology-directed gene repair. These findings opened new perspectives for HR-based genome editing in higher eukaryotes. Here, we demonstrate by using donor DNA templates together with the adeno-associated virus (AAV) Rep78 and Rep68 proteins that sequence- and strand-specific cleavage at a native, predefined, human locus can also greatly enhance homology-directed gene targeting. Our findings argue for the development of other strategies besides direct induction of double-strand chromosomal breaks to achieve efficient and heritable targeted genetic modification of cells and organisms. Finally, harnessing the cellular HR pathway through Rep-mediated nicking expands the range of strategies that make use of AAV elements to bring about stable genetic modification of human cells.     

Proteins encoded by the DpnII restriction gene cassette. Two methylases and an endonuclease

Proteins encoded by three genes in the DpnII restriction enzyme cassette of Streptococcus pneumoniae were purified and characterized. Large amounts of the proteins were produced by subcloning the cassette in an Escherichia coli expression system. All three proteins appear to be dimers composed of identical polypeptide subunits. One is the DpnII endonuclease, and the other two are DNA adenine methylase active at 5' GATC 3' sites. Inactivation of enzyme activity by insertions into the genes and comparison of the DNA sequence with the amino-terminal sequence of amino acid residues in the proteins demonstrated the following correspondence between genes and enzymes. The promoter-proximal gene in the operon, dpnM, encodes a 33 X 10(3) Mr polypeptide that gives rise to a potent DNA methylase. The next gene, dpnA, encodes the 31 x 10(3) Mr polypeptide of a weaker and less-specific methylase. The third gene, dpnB, encodes the 34 x 10(3) Mr polypeptide of the endonuclease. Although the endonuclease polypeptide is initiated from an ordinary ribosome-binding site, each of the methylase polypeptide begins at an atypical site with a consensus sequence entirely different from that of Shine & Dalgarno. This presumptive novel ribosome-binding site is well recognized in both S. pneumoniae and E. coli.     

Mutagenesis of retroviral vectors transducing human beta-globin gene and beta-globin locus control region derivatives results in stable transmission of an active transcriptional structure.

Retrovirus-mediated gene transfer of the human beta-globin gene into hematopoietic stem cells is an attractive approach to the therapy of human beta-globin gene disorders. However, expression of the transduced beta-globin gene linked to its proximal cis-acting sequences (-0.8 to +0.3 kb from the cap site) is considerably below the level required for a significant therapeutic effect. The discovery of the beta-locus control region (beta-LCR), organized in four major DNase I hypersensitive sites far upstream of the human beta-like globin gene cluster, provided a potential means to achieve a high level of expression of a linked human beta-globin gene, but initial attempts to incorporate beta-LCR derivatives in retroviral vectors resulted in the production of low-titer viruses with multiple rearrangements of the transmitted proviral structures. We now describe how extensive mutagenesis of the transduced beta-globin gene, eliminating a 372 bp intronic segment and multiple reverse polyadenylation and splicing signals, increases viral titer significantly and restores stability of proviral transmission upon infection of cell lines and bone marrow-repopulating cells. These optimized vectors have enabled us to analyze the expression properties of various retrovirally transduced beta-LCR derivatives in dimethylsulfoxide-induced murine erythroleukemia cells and to achieve ratios of human beta-globin/murine beta maj-globin mRNA, on a per gene basis, as high as 80%.     

Efficient recombinase-mediated cassette exchange at the AAVS1 locus in human embryonic stem cells using baculoviral vectors

Insertion of a transgene into a defined genomic locus in human embryonic stem cells (hESCs) is crucial in preventing random integration-induced insertional mutagenesis, and can possibly enable persistent transgene expression during hESC expansion and in their differentiated progenies. Here, we employed homologous recombination in hESCs to introduce heterospecific loxP sites into the AAVS1 locus, a site with an open chromatin structure that allows averting transgene silencing phenomena. We then performed Cre recombinase mediated cassette exchange using baculoviral vectors to insert a transgene into the modified AAVS1 locus. Targeting efficiency in the master hESC line with the loxP-docking sites was up to 100%. Expression of the inserted transgene lasted for at least 20 passages during hESC expansion and was retained in differentiated cells derived from the genetically modified hESCs. Thus, this study demonstrates the feasibility of genetic manipulation at the AAVS1 locus with homologous recombination and using viral transduction in hESCs to facilitate recombinase-mediated cassette exchange. The method developed will be useful for repeated gene targeting at a defined locus of the hESC genome.