This is a resubmission of a proposal requesting five years of support to carry out studies aimed at elucidating mechanisms and rules of homologous and nonhomologous recombination in mammalian cells. The applicant utilizes the APRT locus in CHO as his primary experimental system, although the use of other loci and different cell types is planned, when appropriate.
Four specific aims are proposed.
The first aim will examine the influence of repeated DNA sequences and functional elements on recombination. A novel color-based recombination system will be used. In this system, two allelic forms of the green fluorescent protein (GFP) will be used; one green and the other blue due to a single site, although four other silent site variations are also present. Using this system, the speed by which measurements of both homologous and illegitimate recombination occur can be greatly increased. The ease by which the chromosomal location can be modified is also significantly enhanced in the system by utilization of FLP-mediated targeting. Thus, it should be relatively straightforward to not only construct a variety of test situations but also to rapidly measure recombination differences. For example, repeated sequences, such as di- and trinucleotide repeats, can be tested for influences over both homologous and illegitimate recombination. Examination of the influence of transcription on recombination can be tested utilizing the tet activator system which allows tight, regulated control of transcription. Also, the influence of the SV40 origin-of-replication will be tested since it has been shown that the origin alone (in the absence of large T) can promote local instability.
The second aim i s to search for unstable sites within the mammalian genome. While there is a relatively long history of marker integration into mammalian chromosomes, most experience has included the SV40 origin in the incoming construct, which is inherently unstable. Thus, while dogma concludes that the genome is rather unstable, based on these studies, this result is essentially flawed due to the inclusion of the SV40 origin. The applicant plans to integrate a gpt-APRT cassette (without any SV40 sequence), select integrants and measure the stability by frequent loss of both activities. When unstable sites are identified, flanking sequence will be identified and it will be determined if this instability occurs pre- or post-integration. Among the experiments proposed will be to map particularly unstable sites in the human genome, which has the potential to correlate to sites of known instability.
Aim three will investigate means of accomplishing site-specific genome modification. Three tools will be developed. One will be to improve the observation by the applicant that electroporated rare-cutting restriction endonucleases can increase recombination. The plan is to place the coding region for such an enzyme (I-SceI) under control of a regulated promoter, such as tet, which could generate a cell line in which all cells produce the enzyme in a temporal fashion. The second tool to be investigated is the use of triplex-forming oligonucleotides. The plan here is to see if triplex formation can stimulate recombination using a site within APRT which forms relatively stable high affinity triplex formations. Using psoralen-modified oligos one can ask if crosslinking influences recombination. The third tool is RNA/DNA hairpin oligos. This approach takes advantage of the stability of such hairpins and offers much potential for rapid genome modification. Although interesting preliminary results have been obtained, the applicant proposes a careful look at this method using the APRT system and asking relatively simple, but important, questions such as length of the oligos, RNA/DNA ratio and transfection differences.
The final aim will be to investigate recombination in the context of DNA repair. Specifically, using the repair-defective ERCC-1 CHO cell line, the applicant proposes to ask if this cell line behaves differently from the wild-type CHO. Considering what is known about ERCC-1 function, a number of questions can be put forth. For example, the effect of nonhomologous tails could be quite interesting as could be triplex formation under repair-defective states.
