3.

7 Artificial Chromosomes The major limitation of most of the vectors that we have discussed so far is the size limit of the DNA that can be cloned into them. Natural eukaryotic chromosomes consist of hundreds or thousands of genes, together with DNA elements required for chromosomal stability and function such as telomeres and centromeres. Telomeres, which consist of DNA and protein, are located at the ends of chromosomes and protect them from damage. Centromeres are segments of highly repetitive DNA that are essential for the proper control of chromosome distribution during cell division. A logical extension of vector design to clone very large DNA fragments is, therefore, to reconstruct an autonomously replicating chromosome into which DNA fragments may be cloned. Cloning in this way is conceptually similar to cloning in phage with the reconstruction of a replication competent DNA molecule except that the scale of the foreign DNA that can be cloned is much greater.

3.7.1 YACs Yeast artificial chromosome (YAC) vectors allow the cloning, within yeast cells, of fragments of foreign genomic DNA that can approach 500 kbp in size. These vectors contain several elements of typical yeast chromosomes, including the following. A yeast centromere (CEN4). The yeast centromere is specified by a 125 bp DNA segment. The consensus sequence consists of three elements: a 7886 bp region with more than 90 per cent AT residues, flanked by a conserved sequence on one side and a short consensus sequence on the other (reviewed by Clarke (1990)). Yeast autonomously replicating sequence (ARS1). Yeast ARS elements are essentially origins of replication that function in yeast cells autonomously from the replication of yeast chromosomal replication origins. Yeast telomeres (TEL). Telomeres are the specific sequences (5TGTGGGTGTGGTG-3) that are present at the ends of chromosomes in multiple copies and are necessary for replication and chromosome maintenance. Genes for YAC selection in yeast. The vector has a functional copy of URA3, a gene involved in uracil biosynthesis, and TRP1, a gene involved in tryptophan biosynthesis, that allow selection of yeast cells that have taken up the vector. The YAC is transformed into a host yeast cell that is defective in these biosynthetic pathways, and transformants are identified by their ability to complement the nutritional defect.

 Bacterial replication origin and a bacterial selectable marker. In order to propagate the YAC vector in bacterial cells, prior to insertion of genomic DNA, YAC vectors usually contain the ColE1 ori and the ampicillin resistance gene for growth and analysis in E. coli. The cloning of DNA fragments into a YAC is shown diagrammatically in Figure 3.18. The YAC is cleaved using restriction enzymes to generate two arms that each have a telomere sequence at the end. One of the arms contains an autonomous replication sequence (ARS1), a centromere (CEN4) and a

selectable marker (TRP1). The other arm contains a second selectable marker (URA3). Large DNA fragments (>100 kbp) are then ligated between the two arms (Anand, Villasante and Tyler-Smitu, 1989). The insertion of foreign DNA into the cloning site inactivates the suppressor tRNA gene SUP4, expressing tRNATyr, in the vector DNA. In an ade2ochre host yeast cell, the expression of SUP4 results in the formation of white colonies, while in those in which it has been insertionally inactivated will give rise to red yeast colonies (Burke, Carle and Olson, 1987). Yeast cells that are mutated in the ADE2 gene product (coding for the enzyme phosphoribosylamino-imidazole-carboxylase) have a block in the adenine biosynthetic pathway, causing an intermediate to accumulate in the vacuole. This intermediate gives the cell a red colour. The recombinant YACs are therefore transformed into a yeast strain that has defects in its chromosomal copies of the ura3, trp1 and ade2 genes. Transformants are identified as those red colonies that grow on media lacking both uracil and tryptophan. This ensures that the cell has received an artificial chromosome with both telomeres (because of complementation of the two nutritional mutations) and the artificial chromosome contains insert DNA (because the cell is red). There are difficulties associated with working with YACs. Some of these are listed below. Very large DNA molecules are very fragile and prone to breakage, leading to problems of rearrangement. It is estimated that between 10 and 60 per cent of clones in YAC genomic libraries are chimaeric, i.e. regions from different parts of the genome become joined in a single YAC clone (Green et al., 1991). Clones tend to be unstable, with their foreign DNA inserts often being deleted. Naturally occurring repetitive DNA sequences are rare in the yeast genome, and the insertion of such sequences from, say, human DNA inserts appears to increase the recombination frequency within the YAC. This may make the YAC unstable. Interestingly, however, larger YAC vectors are more stable in yeast than shorter ones, which consequently favours cloning of large stretches of DNA (Smith, Smyth and Moir, 1990). There is a high rate of loss of the entire YAC during mitotic growth. It is difficult to separate the YAC from the other host chromosomes because of their similar size. Separation requires sophisticated pulsed-field gel electrophoresis (PFGE). The yield of DNA is not high when the YAC is isolated from yeast cells.

3.7.3 BACs Bacterial artificial chromosomes (BACs) are engineered versions of F plasmids (Shizuya et al., 1992). BACs are capable of carrying approximately 200 kbp of inserted DNA sequence, and the F-factor origin of replication (oriS) maintains their level at approximately one copy per cell. In addition to oriS, BACs contain four F-factor genes required for replication and maintenance of copy number, repE, parA, parB and parC. The overall architecture of a typical BAC is shown in Figure 3.20. In addition to the F-factor genes, pBeloBac11 also contains a selectable antibiotic resistance maker (CAMR) and the lacZ gene harbouring a multiple cloning site for the bluewhite screening of BACs containing inserts (Kim et al., 1996b). Additionally, the BAC contains a cos site (cosN) and a loxP site. These sites are used for specific cleavage of the insert containing BAC during restriction mapping. The cosN site can be cleaved using terminase (Rackwitz et al., 1985), while the loxP site can be cleaved by the Cre protein in the presence of an oligonucleotide to the loxP sequence (Abremski, Hoess and Stanbers, 1983). Additional BACs have been constructed that contain the recognition sites for extremely rare-cutting restriction enzymes. For example, I-SceI is an intron encoded restriction enzyme from the mitochondria of the

 yeast Saccharomyces cerevisiae (Monteilhet et al., 1990). Its large recognition sequence (5-TAGGGATAACAGGGTAAT-3) does not occur once in the human genome, and is consequently very useful for linearizing the vector without cleaving the insert DNA fragments. The DNA inserted into a BAC appears to be very stable. It can survive intact for many hundreds of generations in E. coli cells, and appears to be less prone to rearrangements and deletions when maintained in a recombination defective E. coli host cell. The main drawback of using BAC vectors is that they are present in only one or two copies per cell. This can complicate isolation and screening.

 3.7.4 HACs Human artificial chromosomes (HACs) have been constructed that can survive for extended periods in tissue culture cells (Harrington et al., 1997) (Figure 3.21). As we have already seen, three elements are required for the stability of linear chromosomes centromeres, telomeres and an origin of replication. The human telomere repeat sequence (5-TTAGGG-3) is well known, but it is distinct from its yeast equivalent and the two are not interchangeable. A YAC will not function as a chromosome in human cells. To aid in the isolation of human centromere and replication origin sequences, HACs have been constructed that can be transfected into and maintained within

human cells (Henning et al., 1999). This approach identified multiple repeats of a 171 bp DNA sequence (called an satellite repeat) contained within a 3 million base pair DNA fragment of the human X chromosome that functions as a centromere (Schueler et al., 2001). The repetitive nature of these sequences makes them difficult to study and makes the identification of the centromere itself extremely hard. The human genome contains multiple origins of replication. The average human chromosome contains approximately 150 106 bp of DNA, while DNA polymerase functions maximally at about 3000 replicated bases per minute. Were replication to begin at a single site, each chromosome would take over a month to be replicated, rather than the hour it actually takes. Multiple replication origins mean that there are many places on the eukaryotic chromosome where replication can begin and that the process of complete replication proceeds at a more rapid rate. The sequence of the human replication origin is very degenerate (Vashee et al., 2001), but the development of HACs should allow a more precise mapping of these regions.

 HACs have great potential as tools for both basic research and medical therapy. Artificial chromosomes may ultimately lead to gene therapy vectors (see Chapter 13) with some advantages over existing viral based vectors. They exist as extrachromsomal elements and so would not result in insertional mutagenesis. They should have no size constraint on the amount on DNA that they could carry. By virtue of differences between centromere behaviour in mitosis and meiosis, they might be designed not to function in the germ line. These possibilities remain for the future and will depend on having a much greater understanding of chromosome function than is currently available.