Radiation hybrid mapping

Another powerful method for creating physical maps is the use of irradiation and fusion to gene transfer (IFGT) to produce radiation hybrid maps. This technology originally developed by Goss and Harris, (1975) demonstrated that chromosome fragments, generated by lethal irradiation of donor human cells could be rescued by fusion to rodent recipient. Radiation maps are based on breaks induced by radiation and the resolution of radiation hybrid mapping is a function of both fragment size and retention frequencies, which is the percentage of radiation hybrid cell lines containing a

given chromosome marker. The fragment size can be varied by altering the radiation dose and it is possible to construct panels designed either for map continuity with few markers or for high resolution with large number of markers. Information on localisation using radiation hybrids is obtained by determining linkage to markers of the framework. Such a linkage is expressed by a distance measured in breakage frequency accompanied by a likelihood estimate expressed as a lod score. Distances between markers in the radiation hybrid maps are expressed in cRsooo where 1 cR (n rad) correspond to a 1% frequency o f breakage between two markers after exposure to '"N” rad of X-rays (Cox et a l, 1990; Boehnke a/., 1991).

Cox and co-workers (1990) modified the original approach by using somatic cell hybrid containing a single human chromosome as the donor cell instead of a diploid human cell. This approach is not feasible for mapping an entire genome, as it would require over 4000 hybrids. Walter and co-workers (1994) have found a solution to this problem by reverting to the original protocol of Goss and Harris (1975) and using a diploid human fibroblast as a donor. Using 44 radiation hybrids they constructed a map of the human chromosome 14 containing 400 ordered markers and concluded that a high resolution map of the whole genome is feasible with only a single panel o f 100-200 radiation hybrids.

Radiation hybrid mapping complements both recombination maps and physical maps based on contigs. Whereas recombination maps are constrained by the recombination rate and only limited to polymorphic markers, radiation hybrid methods can be used to map both polymorphic markers and non-polymorphic STSs and ESTs. Therefore a radiation hybrid map o f the human genome was developed to facilitate the completion of the YAC contig of the human genome as it presented an independent means o f integrating STS with the polymorphic markers (Gyapay et a l, 1996). The limitations of YAC contig based maps, for example poor YAC coverage in some regions o f the genome such as terminal parts o f chromosomes and particularly GC-rich regions, which are more likely to be rich in genes, further emphasised the need for an independent method o f mapping. The radiation hybrid map, which consisted o f 168 whole-genome radiation hybrids, was constructed by irradiating donor human fibroblasts at 3000 rads, a relatively low dose to ensure continuity o f the map, and fusing with recipient hamster

cells. A framework map o f 404 markers was constructed by typing the panel with polymorphic markers and ordering them using standard methods (Cox et a/., 1990, Boehnke et a i , 1991). Even though the order o f these markers in the genetic linkage map and the radiation map was the same there were some discrepancies in distances. Some gaps that correspond to 7-10 cM in the genetic map corresponded to very short distances on the radiation map and conversely some clusters on the genetic linkage map were split on the radiation map. Nevertheless the utility o f the map has been demonstrated by mapping o f 374 ESTs, whose accurate localisation on the radiation map was verified by assigning these ESTs to YACs that map to the same interval.

A subset o f 93 hybrids from this panel has been made widely available for genome mapping projects under the panel name Genebridge4. Since the creation o f Genebridge4 panel, which classifies as a lower resolution panel, two other radiation hybrid panels o f higher resolution have been constructed. The Stanford G3 panel is a medium resolution 10,000-rad panel consisting o f 83 clones and the Stanford TNG is a high resolution 50,000-rad panel consisting o f 90 clones, the details o f these panels are available from http ; //shgc-www. stanford. edu.

Ï.5 Physical maps and the current progress of the Human Genome Mapping Project

The physical mapping phase o f the human genome project was officially completed with the landmark publication o f the final YAC contig map o f the human genome, which provided 75% coverage o f the entire human genome (Chumakov et a i ,

1995). This map although a major achievement, still fell short o f the requirement set by the Human Genome Project (HGP) for a ‘sequence ready’ physical map. Therefore, since reaching this goal efforts have been directed at further map construction, integration and validation to meet the final goal set by HGP, which is to produce a physical map o f the human genome containing 30,000 unique markers ordered with respect to each other and spaced on average every 1000,000 base pairs (1 Mb). The physical map constructed by Hudson et ai. (1995) at the Whitehead Institute for Biomedical Research goes more than halfway to meet the criteria set by the human Genome Project. This map includes a large number o f STSs previously generated and mapped by other groups. Approximately 7000

STSs on the map represent genetic markers that have been previously placed on meiotic linkage maps (Gyapay et al., 1994; Dib et al., 1996). Finally STSs developed from random genomic DNA and Expressed Sequence Tags (ESTs) from dbEST database (division of GenBank allocated for expressed sequences) have also been incorporated into the Whitehead physical map. Therefore this map integrates the locations o f a large number o f genes with meiotic linkage maps of the human genome. The two independent means used to order these STSs were radiation hybrid mapping (RH) using the Genebridge4 Radiation Hybrid panel and YAC-STS content mapping. Combining RH and YAC-STS content mapping information Hudson et al. (1995) estimate that they have achieved an effective resolution of 1 Mb and that their map provides RH coverage of 99% o f the genome and physical coverage of 94% o f the genome. Even though the Whitehead map is of immense use to disease gene hunters, in its present form this map does not provide adequate scaffold for sequencing the human genome. Presently, human DNA cloned as bacterial artificial chromosomes (BACs), with an average insert size of

100 kb appears to be the most likely source of template for large scale DNA sequencing. As such a contiguous map containing an ordered marker every 100 kb on average is required to isolate and order efficiently the 100-kb BAC sequencing templates. However as far as ordering of STSs is concerned, physical maps such as the Whitehead map with STSs ordered at an effective resolution of 1 Mb still represent a giant step forward in reaching this ultimate goal. The most up to date physical mapping information on this map can be obtained through the Whitehead Institute for Biomedical Research/MIT Centre for Genome Research’s Human Physical Mapping Project World Wide Web server: fhttp://www-genome.wi.mit.edu/). Meanwhile further characterisation and sequencing of individual chromosomes have already been initiated at various genome centres that took up the task o f constructing ‘sequence ready’ maps of a particular chromosome at the very outset of the human genome project. Information on these

genome centres can be obtained from the World Wide Web server:

(http://webace.sanger.ac.uk/HGPA. The most up to date physical mapping information on these chromosomes is available thorough the electronic genomic databases maintained by the respective genome centres.