Construction of a PAC vector system for the propagation of genomic DNA in bacterial and mammalian cells and subsequent generation of nested deletions in individual library members.

The BAC and PAC cloning systems allow investigators to propagate large genomic DNA fragments up to 300 kb in size in E. colicells. We describe a new PAC shuttle vector that can be propagated in both bacterial and human cells. Specifically, the P1 cloning vector pAd10sacBII was modified by the insertion of a puromycin-resistance gene (pac), the Epstein-Barr Virus (EBV) latent replication origin oriP,and the EBV EBNA1 gene. Transfection studies in HEK 293 cells demonstrated that the modified vector was stably maintained as an episome for at least 30 generations. And since pJCPAC-Mam1 contains a loxP site, genomic DNA cloned into this vector can be subjected to loxP-Cre -mediated deletion events. The transposon vector pTnPGKpuro/loxP was modified to make this system amenable to propagation in human cells by inserting pac, oriP, and EBNA1 elements into the vector (Chatterjee, P.K., Coren, J.C., 1997. Isolating large nested deletions in PACs and BACs by in vivo selection of P1 headful-packaged products of Cre-catalyzed recombination between the loxP site in PAC and BAC and one introduced in transposition. NAR 25, 2205-2212.). pTnPGKpuro/loxP-EBV was then used to generate deletions in an individual library member to demonstrate that all of the deletions still contain the required eukaryotic elements and that they were nested. All library members constructed in pJCPAC-Mam1 can be directly transformed into human cells to assess function. And the deletion technology can be used to aid in delineating the boundaries of genes and other cis-acting elements.

Isolating large nested deletions in bacterial and P1 artificial chromosomes by in vivo P1 packaging of products of Cre-catalysed recombination between the endogenous and a transposed loxP site.

A general approach for isolating large nested deletions in P1 artificial chromosomes (PACs) and bacterial artificial chromosomes (BACs) by retrofitting with a loxP site-containing Tn10 mini-transposon is described. Cre-mediated recombination between the loxP site existing in these clones and one introduced by transposition leads to deletions and inversions of the DNA between these sites. Large deletions are selectively recovered by transducing the retrofitted PAC or BAC clones with P1 phage. The requirement that both loxP sites in the cointegrate be packaged into a P1 head ensures that only large deletions are rescued. PCR analyses identified these deletions as products of legitimate recombination between loxP sites mediated by Cre protein. BACs produce deletions much more efficiently than PACs although the former cannot be induced to greater than unit copy in cells. Mammalian cell-responsive antibiotic resistance markers are introduced as part of the transposon into genomic clone deletions for subsequent functional analysis. Most importantly, the loxP site retrofitting and P1 transduction can be performed in the same bacterial host containing these clones directly isolated from PAC or BAC libraries. These procedures should facilitate physical and functional mapping of genes and regulatory elements in these large plasmids.

Headful packaging revisited: the packaging of more than one DNA molecule into a bacteriophage P1 head.

Like a variety of other bacteriophages, such as T4 and P22, bacteriophage P1 packages DNA by a "headful" mechanism in which the capacity of the viral capsid determines the size of the single DNA molecule that is packaged. Because of the long-standing and general acceptance of this packaging mechanism, we were surprised to discover that some of our observations, using the in vitro P1 packaging system, could be explained by the packaging of less than headful-sized (< 110 kb) DNA molecules into a P1 capsid. To account for these observations, we describe results that support a model of in vitro P1 packaging in which multiple less than headful-sized DNA molecules are taken into a P1 head until that head has been filled. The results further suggest that the phage so generated can occasionally inject more than one DNA molecule into a cell upon viral infection. The data that supports these conclusions are: (1) the DNAs of the circular P1 cloning vectors pAd10sacBII (32 kb) and pNS358 (14 kb) are packaged in vitro with an efficiency of about 6 to 12% of that of longer concatemers of these DNAs. (2) The in vitro packaging of two differentially marked, less than 18 kb plasmid DNAs in the same reaction results in the production of a phage that can occasionally inject both DNAs into the same cell upon infection. (3) Virus particles generated by the packaging of either pAd10sacBII plasmid DNA or the two differently marked plasmids have a density in CsCl equilibrium gradients that is the same as P1 plaque-forming phage, suggesting that the former phage contain a headful of DNA. These results cannot be explained by Cre-mediated site-specific recombination between plasmids in the P1 packaging extracts. Finally, we present in vivo experiments that are also consistent with the headful packaging of multiple DNAs into a P1 head.

Purification of a telomere-binding protein from Physarum polycephalum.

We have purified a telomere-binding protein (PPT) from the acellular slime mold Physarum polycephalum. As shown previously (Coren, J.S., Epstein, E.M. and Vogt, V.M. (1991) Mol. Cell. Biol. 11, 2282-2290), in vitro this protein binds specifically to the double stranded (TTAGGG)n repeats that are found at the telomeres of extrachromosomal ribosomal DNA from this organism, and also at telomeres of mammalian chromosomes. PPT was purified from Physarum nuclear extracts by heat treatment at 90 degrees C followed by heparin-agarose fractionation and gel filtration chromatography. The most purified fraction contained two major protein bands of 10 and 7 kDa when analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. In gel filtration chromatography PPT migrated with a Stokes radius of 1.6 nm. Along with the previously determined sedimentation coefficient of 1.2 S, this value implies a molecular weight of about 8000, making PPT the smallest known telomere-binding protein.

Characterization of a telomere-binding protein from Physarum polycephalum.

We have partially purified a nuclear protein (PPT) from Physarum polycephalum that binds to the extrachromosomal ribosomal DNA telomeres of this acellular slime mold. Binding is specific for the (T2AG3)n telomere repeats, as evidenced by nitrocellulose filter binding assays, by gel mobility shift assays with both DNA fragments and double-stranded oligonucleotides, and by DNase I footprinting. PPT is remarkably heat stable, showing undiminished binding activity after incubation at 90 degrees C. It sediments at 1.2S, corresponding to a molecular weight of about 10,000 (for a globular protein), and its binding activity is undiminished by incubation with RNase, suggesting that it is not a ribonucleoprotein. We hypothesize that PPT plays a structural role in telomeres, perhaps preventing nucleolytic degradation or promoting telomere extension by a telomere-specific terminal transferase.