Mapping of mammalian genomes with radiation (Goss and Harris) hybrids

Comparative study on synteny between yeasts and vertebrates

We studied synteny conservation between 18 yeast species and 13 vertebrate species in order to provide a comparative analysis of the chromosomal plasticity in these 2 phyla. By computing the regions of conserved synteny between all pairwise combinations of species within each group, we show that in vertebrates, the number of conserved synteny blocks exponentially increases along with the divergence between orthologous protein and that concomitantly; the number of genes per block exponentially decreases. The same trends are found in yeasts but only when the mean protein divergence between orthologs remains below 36%. When the average protein divergence exceeds this threshold, the total number of recognizable synteny blocks gradually decreases due to the repeated accumulation of rearrangements. We also show that rearrangement rates are on average 3-fold higher in vertebrates than in yeasts, and are estimated to be of 2 rearrangements/Myr. However, the genome sizes being on average 200 times larger in vertebrates than in yeasts, the normalized rates of chromosome rearrangements (per Mb) are about 50-fold higher in yeast than in vertebrate genomes.

Piggy-BACing the human genome

Using the INRA-Minnesota porcine radiation hybrid panel, we have constructed a human-pig comparative map composed of 2274 loci, including 206 ESTs and 2068 BAC-end sequences, assigned to 34 linkage groups. The average spacing between comparative anchor loci is 1.15 Mb based on human genome sequence coordinates. A total of 51 conserved synteny groups that include 173 conserved segments were identified. This radiation hybrid map has the highest resolution of any porcine map to date and its integration with the porcine linkage map (reported here) will greatly facilitate the positional cloning of genes influencing complex traits of both agricultural and biomedical interest. Additionally, this map will provide a framework for anchoring contigs generated through BAC fingerprinting efforts and assist in the selection of a BAC minimal tiling path and assembly of the first sequence-ready map of the porcine genome.

Conversion technology and its role in genetic testing of inherited diseases

Identification of germline mutations in a sensitive and specific manner presents a continuing challenge. A major contributing factor to this is that humans are diploid and therefore mutations in one allele are often masked by the normal sequence present on the other copy of the chromosome. Mutation analysis on haploid templates (one copy of a chromosome), rather than on diploid templates (both copies of a chromosome), overcomes this problem and obscured mutations are unmasked. Conversion technology converts a sample from diploid to haploid state by isolating individual alleles in somatic cell hybrids. From each sample, a series of stable hybrids is generated that contains the human chromosome complement in the haploid state. This article describes conversion technology and its applications. The utility of this technique in increasing the sensitivity of genetic testing has been demonstrated for the predisposition to hereditary nonpolyposis colorectal cancer and it is proposed that a similar approach may be applicable to many different diseases.

Reading maps of the genes: interpreting the spatiality of genetic knowledge

Genetics has become the pre-eminent interpretation of the body and health and illness. This paper engages with a central technique and metaphor of the new genetics-gene mapping. Through an exploration of the process of gene mapping, the paper argues that the genetic material of the body is spatialised and transformed into a knowable and manipulable entity. Three interpretations of this spatial transformation of the body's materiality are discussed, in turn drawing on Foucault's notion of the construction of medical knowledge, the deconstruction of geographical maps and Haraway's 'fetishised' conception of the gene map. The paper concludes by considering contestations to this dominant discourse, and begins to construct an alternative spatialisation of the body that attempts to 'place' the gene more appropriately in a socially-embedded body and health.

Genetic analysis of inherited hypertension in the rat

Blood pressure is a quantitative trait that has a strong genetic component in humans and rats. Several selectively bred strains of rats with divergent blood pressures serve as an animal model for genetic dissection of the causes of inherited hypertension. The goal is to identify the genetic loci controlling blood pressure, i.e., the so-called quantitative trait loci (QTL). The theoretical basis for such genetic dissection and recent progress in understanding genetic hypertension are reviewed. The usual paradigm is to produce segregating populations derived from a hypertensive and normotensive strain and to seek linkage of blood pressure to genetic markers using recently developed statistical techniques for QTL analysis. This has yielded candidate QTL regions on almost every rat chromosome, and also some interactions between QTL have been defined. These statistically defined QTL regions are much too large to practice positional cloning to identify the genes involved. Most investigators are, therefore, fine mapping the QTL using congenic strains to substitute small segments of chromosome from one strain into another. Although impressive progress has been made, this process is slow due to the extensive breeding that is required. At this point, no blood pressure QTL have met stringent criteria for identification, but this should be an attainable goal given the recently developed genomic resources for the rat. Similar experiments are ongoing to look for genes that influence cardiac hypertrophy, stroke, and renal failure and that are independent of the genes for hypertension.

Estimation of distances and map construction using radiation hybrids

A method of estimating distances between pairs of genetic markers is described that directly uses their observed joint frequency distribution in a panel of radiation hybrids (RHs). The distance measure is based on the strength of association between marker pairs, which is high for close markers and decays with distance. These distances are then submitted to a previous method that generates linear coordinates for the markers directly from the intermarker distance matrix. This method of map building from RH data is simpler than others, because it uses only the observed joint frequency distributions of markers in the panel, and does not attempt to model unobserved quantities such as the retention of different sized fragments that contain the markers. It also incorporates directly the observed variation in retention of different markers, without needing a model for differential fragment retention dependent on chromosomal location, which is generally not known. Only small, precise distances are used in map construction, thereby reducing any effects of different fragment retention frequencies and local variations in X-ray sensitivity. The method is tested by simulation, and known marker distances and locations are successfully recovered from RH raw data. The method is also applied to publicly available data sets related to the recent transcript map of the human genome.

A first-generation whole genome-radiation hybrid map spanning the mouse genome

We have assembled a first-generation anchor map of the mouse genome using a panel of 94 whole-genome-radiation hybrids (WG-RHs) and 271 sequence-tagged sites (STSs). This is the first genome-wide RH anchor map of a model organism. All of the STSs have been previously localized on the genetic map and are located 8.8 Mb apart on average. This mouse WG-RH panel, known as T31, has an average retention frequency of 27.6% and an estimated potential resolution of 145 kb, making it a powerful resource for efficient large-scale expressed sequence tag mapping. [All of the mapping data for the maps presented here have been deposited at the Research Genetics, Inc., web site and can be freely accessed and downloaded at http://www.resgen.com/.]

Identification of a novel breast-cancer-anti-estrogen-resistance (BCAR2) locus by cell-fusion-mediated gene transfer in human breast-cancer cells

Development of anti-estrogen resistance limits the benefit of endocrine therapy of breast cancer. The mechanistic basis for resistance to the anti-estrogen tamoxifen may involve (epi)genetic alterations within tumor cells. We have initiated a random search for genes allowing estrogen-dependent ZR-75-1 human breast-cancer cells to proliferate in the presence of tamoxifen. The strategy was based on insertion mutagenesis of ZR-75-1 cells using defective retrovirus and subsequent identification of common integration sites. As an alternative approach to identify integration loci involved in anti-estrogen resistance, we have applied cell fusion. Integration regions from lethally irradiated, tamoxifen-resistant cells were transferred to hygromycin B-resistant ZR-75-1 cells. Somatic cell hybrids were established by selection for resistance to G418 (encoded by the integrated virus) and hygromycin B. Individual integration loci were thus separated among different cell hybrids and tested for their role in anti-estrogen resistance. Analysis of a panel of 29 somatic-cell hybrids revealed that tamoxifen resistance co-segregated with only 1 of the 2 integration loci present in the tamoxifen-resistant donor cell line. This locus was further identified as a common integration site in our panel of tamoxifen-resistant cell clones. Our results designate this integration site as the second breast-cancer-anti-estrogen-resistance locus (BCAR2), which most likely contains a gene responsible for the anti-estrogen-resistant phenotype in close proximity to the integrated virus. Our data also imply that individual genes can alter the estrogen dependency of human breast-cancer cells in a dominant manner in vitro.

Estimating physical distances from radiation hybrid mapping data

Radiation hybrid mapping has become an established tool for building physical maps. It represents a powerful way of constructing YAC contigs and high-resolution maps for positional cloning experiments. Ideally, radiation hybrids should not only provide support for the true order of the markers, but also accurate estimates of the physical distances between them. Statistical analysis of radiation hybrids has proved difficult because of the number of parameters (representing the fragment retention probabilities) that must be estimated, and simplifying assumptions are needed to analyze large numbers of markers simultaneously. The ramifications of these assumptions for the calculation of physical distances are investigated. A simple two-locus model is presented to demonstrate that variation in marker retention can lead to distortions in the estimates of distance. Multilocus simulations show that, when marker retention is constant across the chromosome, good estimates of physical distance can be derived using simple models of retention. However, further simulations exploring variable retention schemes demonstrate that significant errors in the estimates of map distances can occur. Ways of minimizing these distortions are discussed.

Studies on locus content mapping

Locus content maps are derived from monosomic or disomic chromosomes broken by radiation, shearing, or other clastogen, the fragments being distributed among clones by dilution or incorporation into the cells of another species and scored for segregation of markers. Locus content maps provide evidence about radiosensitivity of chromosome regions, support for order, and approximate location. Omission of the most aberrant and least informative clones increases efficiency of localization. Correct analysis must allow for preferential retention of certain sequences, monosomy or polysomy of donor chromosomes, and error filtration. Combination of these refinements extracts substantially more information from fewer clones. Because of unmodeled peculiarities in the data, the best analysis does not recover the physical map but roughly localizes markers that may be monomorphic and therefore unsuitable for linkage mapping. As with linkage for polymorphic loci, distance in the composite map should be confirmed by physical methods.

Hybrid selection as a method of increasing mapping power for radiation hybrids

Radiation hybrids have become a widely used tool for physical mapping. A drawback of the technique is that large numbers of hybrids are required to construct robust, high-resolution maps. The information contained within a panel of radiation hybrids is limited by the frequency of retention of chromosomal fragments from the donor cell line. In almost all experiments to date, the retention frequency has been below the optimal level; therefore, many hybrids are needed to produce high-quality maps. Because of the labor-intensive nature of large-scale mapping projects, it is important to make panels as small as possible. One method that has been adopted is to produce initially a large number of hybrids that are all typed with a few loci. Those hybrids showing satisfactorily high retention are admitted to the final panel and the rest are discarded. In this way, a panel of radiation hybrids with higher than expected retention can be created. Methods for conducting such a selection regime are discussed. To investigate the potential advantages of selecting hybrids based on their retention frequency, simulations were run under a variety of conditions. As expected panels with high retention (40%) provided better mapping resources than panels with lower (20%) retention. Beginning with an initial panel of 200 hybrids, comparisons of a random sample of 100 hybrids and the set of those 100 hybrids showing the highest marker retention demonstrated that selection may not be always the best strategy despite the increase in mean retention it yields. The selection of hybrids containing large numbers of fragments leads to an overestimation of the frequency of radiation-induced breaks. When breaks occur with high frequency (for example, when high radiation doses are used), the selection of hybrids leads to a loss of linkage and hence an inability to order the markers. As such, the merits of screening hybrids depends on both the radiation dose and the desired map resolution.

Pairwise analysis of radiation hybrid mapping data

Two point lod scores are widely used in pedigree analysis as they provide a fast and efficient method of establishing linkage. Groups of markers that lie in close proximity to one another can be formed by admitting any locus that is linked to at least one existing member of the group with lod score greater than some predetermined value. It seems natural to extend this technique to Radiation Hybrid Mapping both for constructing groups of tightly linked loci that may then be analysed using more powerful statistics and as a method of ordering in its own right. A general extension of two point analysis is derived and the problems associated with radiation hybrid data are discussed. In particular, the additional parameters representing the probabilities of different fragments being retained (which have no parallel in classical linkage analysis) lead to a range of estimators of the breakage probability, O, which have equal and maximal likelihood. Ways of circumventing this problem are discussed along with the potential errors they introduce. Importantly the ambiguity in estimation of theta is not carried through to the lod score as this depends only on the maximum value of the likelihood and not on the particular value of theta at which it occurs. Thus even though two point analysis fails to provide robust estimates of either breakage probabilities or the distance between loci, it represents a simple and effective method of constructing linkage groups that may be analysed with more powerful statistical methods. This is particularly important given the large number of microsatellites, ESTs and candidate genes currently being typed on radiation hybrids.