Patterns of integration of DNA microinjected into cultured mammalian cells: evidence for homologous recombination between injected plasmid DNA molecules.

We examined the fate of DNA microinjected into nuclei of cultured mammalian cells. The sequence composition and the physical form of the vector carrying the selectable gene affected the efficiency of DNA-mediated transformation. Introduction of sequences near the simian virus 40 origin of DNA replication or in the long terminal repeat of avian sarcoma provirus into a recombinant plasmid containing the herpes simplex virus thymidine kinase gene. (pBR322/HSV-tk) enhanced the frequency of transformation of LMtk- and RAT-2tk- cells to the TK+ phenotype 20- to 40-fold. In cells receiving injections of only a few plasmid DNA molecules, the transformation frequency was 40-fold higher after injection of linear molecules than after injection of supercoiled molecules. By controlling the number of gene copies injected into a recipient cell, we could obtain transformants containing a single copy or as many as 50 to 100 copies of the selectable gene. Multiple copies of the transforming gene were not scattered throughout the host genome but were integrated as a concatemer at one or a very few sites in the host chromosome. Independent transformants contained the donated genes in different chromosomes. The orientation of the gene copies within the concatemer was not random; rather, the copies were organized as tandem head-to-tail arrays. By analyzing transformants obtained by coinjecting two vectors which were identical except that in one a portion of the vector was inverted, we were able to conclude that the head-to-tail concatemers were generated predominantly by homologous recombination. Surprisingly, these head-to-tail concatemers were found in transformants obtained by injecting either supercoiled or linear plasmid DNA. Even though we demonstrated that cultured mammalian cells contain the enzymes for ligating two DNA molecules very efficiently irrespective of the sequences or topology at their ends, we found that even linear plasmid DNA was recruited into the concatemer by homologous recombination.     

Molecular cloning and sequence of the concatemer junction from vaccinia virus replicative DNA. Viral nuclease cleavage sites in cruciform structures

The concatemer junction from replicative forms of vaccinia virus DNA was cloned into plasmid vectors and shown to be a precise duplex copy of the viral terminal hairpin structure, with each strand corresponding to one of the alternative sequence isomers. The plasmids were relaxed circles with extruded cruciforms representing two copies of the vaccinia telomere hairpin structure. Head-to-head dimers containing two copies of the vaccinia virus concatemer junction were observed to contain only one set of stem-loop structures per molecule, suggesting that the initial formation of a small cruciform, and not branch migration, was the rate-limiting step in cruciform formation. The plasmids containing the concatemer junction were converted into nicked circular, linear and cross-linked linear molecules by a nuclease isolated from vaccinia virions. The region-specific cleavage near the border of the hairpin loop and the formation of DNA cross-links in some of the molecules is consistent with the nuclease acting as a nicking-closing enzyme that participates in the resolution of mature termini from replicative concatemer intermediates.     

Introduction of stable high-copy-number DNA into Chinese hamster ovary cells by electroporation

A new gene transfer protocol has been developed that introduces up to 800 copies of an expression vector into Chinese hamster ovary cells in a single step by electroporation. The DNA typically integrates in tandem repeats so that the restriction endonuclease site used to linearize the input DNA remains intact. This is likely due to ligation of vector DNA via cohesive ends prior to integration. This high-copy-number procedure is far more rapid than the conventional stepwise gene amplification method used to generate stable eukaryotic protein production cell lines. By employing the expression vector pJODtPA, in which the selectable marker dihydrofolate reductase (DHFR) and the human tissue plasminogen activator (tPA) casettes are separated by a spacer and an RNA polymerase II terminator, cell lines secreting as much as 24 pg/cell.day tPA were isolated following electroporation and a single methotrexate selection. Gene copies and expression levels are stable over long periods of growth. A single round of gene amplification was performed following the high-copy-number procedure to yield a clone having a tPA production level of 45 pg/cell.day.     

Denaturation mapping of R factor deoxyribonucleic acid.

The R factor NR1 consists of two components: a resistance transfer factor which harbors the tetracycline resistance genes (RTF-TC) and the r-determinants component which harbors the other drug resistance genes. Using partial denaturation mapping it is possible to distinguish the RTF-TC region from the r-determinants region of the composite R factor NR1 DNA which has a contour length of 37 mum and a density of 1.712 g/ml. The r-determinants region was a relatively undenatured 8.5-mum segment of the molecule when the deoxyribonucleic acid was partially denatured at pH 10.7. An RTF-TC genetic segregant of NR1 which had lost the r-determinants component had a contour length of 28.7 mum and a density of 1.710 g/ml. Characterization of an RTF-TC using partial denaturation mapping at pH 10.7 confirmed that the relatively undenatured 8.5-mum r-determinants segment of the composite R factor had been deleted. Circular, transitioned NR1 DNA molecules (1.716 to 1.718 g/ml), whose contour lengths were consistent with an RTF-TC plus an integral number of tandem copies of r-determinants, were also characterized by denaturation mapping. The relatively undenatured region in these molecules had a length equal to an integral number of copies of r-determinants and was located at the same site in the partially denatured RTF-TC as the single copy of r-determinants in the 37-mum composite NR1. This indicates that there is a unique integration site for r-determinants in the RTF-TC component. The R factor UCR122, a TC deletion mutant of NR1, was also characterized by denaturation mapping. The translocation of the TC resistance gene(s) on the denaturation map permitted the alignment of the denaturation map with the heteroduplex map of Sharp et al. (u073). Linear and circular monomeric and presumed multimeric r-determinants DNA molecules (p = 1.718 g/ml) were partially denatured at a higher pH (11.10). The r-determinants multimers showed a repeating 8.3-mum (monomeric) partial denaturation pattern indicating a head-to-tail arrangement of monomers in these poly-r-determinant molecules.     

Lack of site specific recombination of exogenous DNA in mouse L cells

Plasmids were constructed containing the HSV thymidine kinase gene and two copies of X. borealis 5S rDNA. Mouse L TK- cells were transformed with these DNAs, with selection for the TK+ gene. Transformed cells were then analyzed by Southern blot hybridization and hybridization in situ to determine whether integration of the exogenous DNA occurred at regions of chromosomal homology i.e., at the 5S rDNA regions. Four cell lines were analyzed by Southern blots. Differences in restriction endonuclease specificity strongly suggested that integration was at a different site in each cell line. Two cell lines were further analyzed by hybridization in situ; each showed a single integration site, both different from each other and different from the mouse L cell 5S rDNA sites. Therefore, the presence of two copies of the 5S rDNA gene in the DNA introduced by gene transfer and approximately 300-350 copies of the mouse 5S rDNA gene was not sufficient in these experiments to produce homologous integration into a specific site.     

Insertional mutagenesis of Aspergillus fumigatus

We have investigated transformation with heterologous DNA as a method for insertional mutagenesis of Aspergillus fumigatus. Two methods, polyethylene glycol-mediated transformation of protoplasts and electroporation of germinating spores, were used to establish conditions leading to single-copy integration of transforming DNA at different genomic sites. We have assessed the effect of restriction enzyme-mediated integration (REMI) for both methods. Non-REMI protoplast transformation led to integration of multiple copies of transforming DNA in the majority of transformants. Results of REMI with protoplast transformation varied depending on the enzyme used. Low concentrations of several restriction enzymes stimulated transformation, but of ten enzymes investigated only REMI with XhoI and KpnI resulted in single-copy integration of transforming DNA for the majority of transformants. For protoplast transformation with XhoI- or KpnI-based REMI, 50% and 76% of insertions, respectively, were due to integrations at a genomic enzyme site corresponding to the enzyme used for REMI. Electroporation of spores without addition of restriction enzyme resulted in a high transformation efficiency, with up to 67% of transformants containing a single copy of transforming DNA. In contrast to protoplast transformation, electroporation of spores in the presence of a restriction enzyme did not improve transformation efficiency or lead to insertion at genomic restriction sites. Southern analysis indicated that for both protoplast transformation with REMI using KpnI or XhoI and for electroporation of spores without addition of restriction enzymes, transforming DNA inserted at different genomic sites in a high proportion of transformants.     

Bacteriophage T7 DNA packaging. III. A "hairpin" end formed on T7 concatemers may be an intermediate in the processing reaction

An unusual left end (M-end) has been identified on bacteriophage T7 DNA isolated from T7-infected cells. This end has a "hairpin" structure and is formed at a short inverted repeat sequence centered around nucleotide 39,587 of T7, 190 base-pairs to the left of the site where a mature left end is formed on the T7 concatemer. We do not detect the companion right end that would be formed if the M-end is produced by a double-stranded cut on the T7 concatemer. This suggests that the hairpin left end may be generated from a single-stranded cut in the DNA that is used to prime rightward DNA synthesis. The formation of M-end does not require the products of T7 genes 10, 18 or 19, proteins that are essential for the formation of mature T7 ends. During infection with a T7 gene 3 (endonuclease) mutant, phage DNA synthesis is reduced and the concatemers are not processed into unit length DNA molecules, but both M-end and the mature right end are formed on the concatemer DNA. These two ends are also found associated with the large, rapidly sedimenting concatemers formed during a normal T7 infection while the mature left end is present only on unit length T7 DNA molecules. We propose that DNA replication primed from the hairpin end produced by a nick in the inverted repeat sequence provides a mechanism to duplicate the terminal repeat before DNA packaging. Packaging is initiated with the formation of a mature right end on the branched concatemer and, as the phage head is filled, the T7 gene 3 endonuclease may be required to trim the replication forks from the DNA. Concatemer processing is completed by the removal of the 190 base-pair hairpin end to produce the mature left end.     

Insertional mutagenesis of Aspergillus fumigatus

We have investigated transformation with heterologous DNA as a method for insertional mutagenesis of Aspergillus fumigatus. Two methods, polyethylene glycol-mediated transformation of protoplasts and electroporation of germinating spores, were used to establish conditions leading to single-copy integration of transforming DNA at different genomic sites. We have assessed the effect of restriction enzyme-mediated integration (REMI) for both methods. Non-REMI protoplast transformation led to integration of multiple copies of transforming DNA in the majority of transformants. Results of REMI with protoplast transformation varied depending on the enzyme used. Low concentrations of several restriction enzymes stimulated transformation, but of ten enzymes investigated only REMI with XhoI and KpnI resulted in single-copy integration of transforming DNA for the majority of transformants. For protoplast transformation with XhoI- or KpnI-based REMI, 50% and 76% of insertions, respectively, were due to integrations at a genomic enzyme site corresponding to the enzyme used for REMI. Electroporation of spores without addition of restriction enzyme resulted in a high transformation efficiency, with up to 67% of transformants containing a single copy of transforming DNA. In contrast to protoplast transformation, electroporation of spores in the presence of a restriction enzyme did not improve transformation efficiency or lead to insertion at genomic restriction sites. Southern analysis indicated that for both protoplast transformation with REMI using KpnI or XhoI and for electroporation of spores without addition of restriction enzymes, transforming DNA inserted at different genomic sites in a high proportion of transformants. Received: 6 March 1998 / Accepted: 25 May 1998     

Viral DNA in bursal lymphomas induced by avian leukosis viruses.

Avian leukosis viruses (ALV) induce malignant lymphoma of the bursa of Fabricius. Viral DNA in tumors and normal tissues from infected birds were analyzed by using restriction endonucleases. Viral DNA fragments diagnostic of the exogenous ALV were easily detected in tumors, uninvolved bursal tissue, kidney, and erythrocyte nuclei. Exogenous viral DNA was more difficult to detect in liver. Using a restriction endonuclease (SacI) which cleaves linear unintegrated ALV DNA in a single site to define integration sites in DNA from the various tissues, we were able to detect ALV DNA only in tumor tissue. We concluded that the proviral DNA detected in the various nontumor tissue must be integrated in multiple sites. The appearance of ALV integration sites uniquely in tumors suggests that they are clonal growths. Furthermore, the data suggested the presence of a single exogenous integration site for the ALV provirus in each of six early neoplastic bursal nodules. This provirus appeared to retain the organization of EcoRI and BamHI recognition sequences present in the genome of virus used to infect the birds. The ALV integration site appeared different in each of the tumors studied. In a widespread metastatic lymphoma, multiple ALV integration sites were found as well as structural alterations in at least some copies of the ALV provirus.     

Analysis of cellular integration sites in avian sarcoma virus infected duck embryo cells.

The cellular sites of integration of avian sarcoma virus (ASV) have been examined in clones of duck embryo cells infected with the Bratislava 77 strain of ASV using restriction endonuclease digestion, agarose gel electrophoresis, Southern blotting, and hybridization with labeled ASV complementary DNA probes. DNA prepared from 11 clones of duck embryo cells infected with the Bratislava 77 strain of ASV was digested with the restriction enzymes HpaI, which cleaves once within the viral genome, and Hind III, which cleaves twice within the viral genome, and the virus-cell DNA juncture fragments were resolved by agarose gel electrophoresis. Analysis of the virus-cell junctures present in individual ASV-infected duck embryo clones revealed that all clones contain at least one copy of nondefective proviral DNA with some clones containing as many as 5 to 6 copies of proviral DNA. A comparison of the virus-cell juncture fragments present in different ASV-infected clones showed that each clone contains a unique set of virus-cell junctures. These data suggest that ASV DNA can integrate at multiple sites within the duck embryo cell genome and that these sites appear to be different as defined by digestion with the restriction enzymes HpaI and HindIII.     

Induction of multiple plasmid recombination in Saccharomyces cerevisiae by psoralen reaction and double strand breaks.

DNA damage-induced multiple recombination was studied by cotransforming yeast cells with pairs of nonreplicating plasmids carrying different genetic markers. Reaction of one of the plasmids with the interstrand crosslinking agent, psoralen, stimulated cellular transformation by the undamaged plasmid. The cotransformants carried copies of both plasmids cointegrated in tandem arrays at chromosomal sites homologous to either the damaged or the undamaged DNA. Plasmid linearization, by restriction endonuclease digestion, was also found to stimulate the cointegration of unmodified plasmids. Disruption of the RAD1 gene reduced the psoralen damage-induced cotransformation of intact plasmid, but had no effect on the stimulation by double strand breaks. Placement of the double strand breaks within yeast genes produced cointegration only at sequences homologous to the damaged plasmids, while digestion within vector sequences produced integration at chromosomal sites homologous to either the damaged or the undamaged plasmid molecules. These observations suggest a model for multiple recombination events in which an initial exchange occurs between the damaged DNA and homologous sequences on an undamaged molecule. Linked sequences on the undamaged molecule up to 870 base pairs distant from the break site participate in subsequent exchanges with other intact DNA molecules. These events result in recombinants produced by reciprocal exchange between three or more DNA molecules.     

Concerted action at eight phosphodiester bonds by the BcgI restriction endonuclease

The BcgI endonuclease exemplifies a subset of restriction enzymes, the Type IIB class, which make two double-strand breaks (DSBs) at each copy of their recognition sequence, one either side of the site, to excise the sequence from the remainder of the DNA. In this study, we show that BcgI is essentially inactive when bound to a single site and that to cleave a DNA with one copy of its recognition sequence, it has to act in trans, bridging two separate DNA molecules. We also show that BcgI makes the two DSBs at an individual site in a highly concerted manner. Intermediates cut on one side of the site do not accumulate during the course of the reaction: instead, the DNA is converted straight to the final products cut on both sides. On DNA with two sites, BcgI bridges the sites in cis and then generally proceeds to cut both strands on both sides of both sites without leaving the DNA. The BcgI restriction enzyme can thus excise two DNA segments together, by cleaving eight phosphodiester bonds within a single-DNA binding event.     

Site-specific integration of Agrobacterium tumefaciens T-DNA via double-stranded intermediates

Agrobacterium tumefaciens-mediated genetic transformation involves transfer of a single-stranded T-DNA molecule (T strand) into the host cell, followed by its integration into the plant genome. The molecular mechanism of T-DNA integration, the culmination point of the entire transformation process, remains largely obscure. Here, we studied the roles of double-stranded breaks (DSBs) and double-stranded T-DNA intermediates in the integration process. We produced transgenic tobacco (Nicotiana tabacum) plants carrying an I-SceI endonuclease recognition site that, upon cleavage with I-SceI, generates DSB. Then, we retransformed these plants with two A. tumefaciens strains: one that allows transient expression of I-SceI to induce DSB and the other that carries a T-DNA with the I-SceI site and an integration selection marker. Integration of this latter T-DNA as full-length and I-SceI-digested molecules into the DSB site was analyzed in the resulting plants. Of 620 transgenic plants, 16 plants integrated T-DNA into DSB at their I-SceI sites; because DSB induces DNA repair, these results suggest that the invading T-DNA molecules target to the DNA repair sites for integration. Furthermore, of these 16 plants, seven plants incorporated T-DNA digested with I-SceI, which cleaves only double-stranded DNA. Thus, T-strand molecules can be converted into double-stranded intermediates before their integration into the DSB sites within the host cell genome.     

Site-specific recombination at oriT of plasmid R1162 in the absence of conjugative transfer.

R1162 is efficiently comobilized during conjugative transfer of the self-transmissible plasmid R751. Bacteriophage M13 derivatives that contain two directly repeated copies of oriT, the site on R1162 DNA required in cis for mobilization, were constructed. Phage DNA molecules underwent recombination during infection of Escherichia coli, with the product retaining a single functional copy of oriT. Recombination was strand specific and depended on R1162 gene products involved in mobilization, but did not require the self-transmissible plasmid vector. Two genes were identified, one essential for recombination and the other affecting the frequency of recombination. Recombination of bacteriophage DNA could form the basis of a simple model for some of the events occurring during conjugation without the complexity of a true mating system.     

Processing of concatemers of bacteriophage T7 DNA in vitro

The T7 chromosome is a double-stranded linear DNA molecule flanked by direct terminal repeats or so-called terminal redundancies. Late in infection bacteriophage T7 DNA accumulates in the form of concatemers, molecules that are comprised of T7 chromosomes joined in a head to tail arrangement through shared terminal redundancies. To elucidate the molecular mechanisms of concatemer processing, we have developed extracts that process concatemeric DNA. The in vitro system consists of an extract of phage T7-infected cells that provides all T7 gene products and minimal levels of endogenous concatemeric DNA. Processing is analyzed using a linear 32P-labeled substrate containing the concatemeric joint. T7 gene products required for in vitro processing can be divided into two groups; one group is essential for concatemer processing, and the other is required for the production of full length left-hand ends. The products of genes 8 (prohead protein), 9 (scaffolding protein), and 19 (DNA maturation) along with gene 18 protein are essential, indicating that capsids are required for processing. In extracts lacking one or more of the products of genes 2 (Escherichia coli RNA polymerase inhibitor), 5 (DNA polymerase), and 6 (exonuclease), full length right-hand ends are produced. However, the left-hand ends produced are truncated, lacking at least 160 base pairs, the length of the terminal redundancy. Gene 3 endonuclease, required for concatemer processing in vivo, is not required in this system. Both the full length left- and right-hand ends produced by the processing reaction are protected from DNase I digestion, suggesting that processing of the concatemeric joint substrate is accompanied by packaging.     

EcoRI analysis of bacteriophage P22 DNA packaging.

Bacteriophage P22 linear DNA molecules are a set of circularly permuted sequences with ends located in a limited region of the physical map. This mature form of the viral chromosome is cut in headful lengths from a concatemeric precursor during DNA encapsulation. Packaging of P22 DNA begins at a specific site, which we have termed pac, and then proceeds sequentially to cut lengths of DNA slightly longer than one complete set of P22 genes (Tye et al., 1974b). The sites of DNA maturation events have been located on the physical map of EcoRI cleavage sites in P22 DNA. EcoRI digestion products of mature P22 wild-type DNA were compared with EcoRI fragments of two deletion and two insertion mutant DNAs. These mutations decrease or increase the length of the genome, but do not alter the DNA encapsulation mechanism. Thus the position of mature molecular ends relative to EcoRI restriction sites is different in each mutant, and comparison of the digests shows which fragments come from the ends of linear molecules. From the positions of the ends of molecules processed in sequential headfuls, the location of pac and the direction of encapsulation relative to the P22 map were deduced. The pac site lies in EcoRI fragment A, 4.1 × 10³ base-pairs from EcoRI cleavage site 1. Sequential packaging of the concatemer is initiated at pac and proceeds in the counterclockwise direction relative to the circular map of P22. One-third of the linears in a population are cut from the concatemer at pac, and most packaging sequences do not extend beyond four headfuls.Fragment D is produced by EcoRI cleavage at a site near the end of a linear chromosome which has been encapsulated starting at pac. The position of the pac site is therefore defined by one end of fragment D. The pac site is not located near genes 12 and 18, the only known site for initiation of P22 DNA replication, but lies among late genes at a position on the physical gene map approximately analogous to the cohesive end site (cos) of bacteriophage λ at which λ DNA is cleaved during encapsulation. Our results suggest that P22 and λ DNA maturation mechanisms have many common properties.     

Properties of electroporation-mediated DNA transfer in Escherichia coli

Efficient and reproducible DNA-transfection was attained in E. coli, by electroporation. The yield of the transfectants was affected by pretreatment of the recipient cells as well as by the composition of the electroporation medium. Using a single pulse procedure, relationships among the electrical parameters, the transfection efficiency, and the cellular viability were investigated in 10 mM Tris-HCl buffer (pH 7.5) containing 5% sucrose. Certain sodium salts (e.g., citrate, phosphate, and sulfate) were promotive, whereas Mg2+, DEAE-dextran, and polyvinylpyrrolidone were inhibitory to the transfection. Heterologous nucleic acids (native DNA, denatured DNA, and tRNA) exerted only a marginal effect on transfection with a viral replicative-form DNA. The efficiency of DNA transfer was affected by culture conditions, and bacteria grown at a higher temperature were more competent. The electroporation system was more efficient than an improved CaCl2 method, not only in transfection with viral single- and double-stranded DNAs, but also in transformation with plasmid DNAs.     

Conjugal mobilization of plasmid DNA: termination frequency at the origin of transfer of plasmid R1162.

Conjugal transfer of plasmid R1162 is initiated and terminated at a 38-bp origin of transfer (oriT). Plasmids containing two directly repeated copies of oriT were used to determine the actual frequency of termination at this site. This frequency was calculated both for oriTnic, a mutated origin that cannot act as the initiation site of transfer, and for an unmutated oriT where transfer had been initiated. In both cases, the termination frequency decreased as the distance between the initiation and termination sites became greater and was significantly less than one for plasmids the size of R1162. A substantial proportion of recipient cells received more than one plasmid copy during transfer. Our results indicate that termination is inefficient but that this is partly compensated for by the transmission of multiple plasmid copies.     

Circular Herpesvirus sylvilagus DNA in spleen cells of experimentally infected cottontail rabbits.

Cottontail rabbits (Sylvilagus floridanus) were infected with Herpesvirus sylvilagus, and spleen cells were analyzed for the presence of virus-specific, covalently closed circular, and linear DNA molecules by a simple electrophoretic technique, followed by transfer to nitrocellulose filters and hybridization with cloned viral DNA (Gardella et al., J. Virol. 50:248-254, 1984). Approximately 0.2 copies per cell of circular DNA and 0.2 copies per cell of linear DNA were detected by hybridization with a cloned viral DNA fragment. The size of the viral DNA was estimated at ca. 158 kilobase pairs. Restriction endonuclease patterns suggested structural similarities to cottontail herpesvirus DNA.