Peter Demant

The nature and identification of quantitative trait loci: a community’s view

This white paper by eighty members of the Complex Trait Consortium presents a communitys view on the approaches and statistical analyses that are needed for the identification of genetic loci that determine quantitative traits. Quantitative trait loci (QTLs) can be identified in several ways, but is there a definitive test of whether a candidate locus actually corresponds to a specific QTL?

Recombinant congenic strains — A new tool for analyzing genetic traits determined by more than one gene

Discovery of the H-2 system by Gorer (1936) and its further analysis (for review of the early work, see Klein 1975, Snell et al. 1976, Festenstein and Demant 1978) led to several unexpected findings concerning the biological effects of the histocompatibility genes or genes closely linked to the major histocompatibility complex (MHC). One of these findings was that the MHC genes in the mouse influence susceptibility to virally induced leukemia (Lilly et al. 1964); another finding was that susceptibility to a number of diseases is specifically associated with certain HLA haplotypes (for review, see Tiwari and Terasaki 1985). The control of these phenomena frequently involves, in addition to the MHC, several genes not linked to the MHC. Our interest in understanding the genetic context in which the action of the MHC genes takes place led us to evaluate the available genetic approaches used to analyze such traits controlled by multiple genes. We show here that in order to achieve sufficient resolution power to analyze such traits, a new genetic system, the recombinant congenic strains (RCS), is needed. We describe the main features of this new system and its advantages and disadvantages as compared with other genetic systems. This tool for genetic analysis will be discussed with respect to its application to tumor biology; however, it can be applied also to other types of multigenic quantitative polymorphisms.

Positional cloning of the combined hyperlipidemia gene Hyplip1.

We have developed technologies that simplify genomic library construction and screening, substantially reducing both the time and the cost associated with traditional library screening methods and facilitating the generation of gene-targeting constructs. By taking advantage of homologous recombination in Escherichia coli, we were able to use as little as 80 bp of total sequence homology to screen for a specific gene from a genomic library in plasmid or phage form. This method, called recombination cloning (REC), takes only a few days instead of the several weeks required for traditional plaque-lift methods. In addition, because every clone in the mouse genomic library we have constructed has a negative selection marker adjacent to the genomic insert, REC screening can generate gene-targeting vectors in one step, from library screening to finished construct. Conditional targeting constructs can be generated easily with minimal additional manipulation.

Cancer susceptibility in the mouse: genetics, biology and implications for human cancer

Growing evidence that a large proportion of apparently non-hereditary sporadic cancers occur in genetically predisposed individuals has emphasized the need to identify the underlying susceptibility genes. Increasingly, it seems that the best approach to define the numerous genes that have small but cumulative effects is to first identify and map them in mice, and subsequently to study the role of their homologues in humans. Development of new gene-mapping resources and strategies in mice has, for the first time, allowed some of these genes to be identified. In future, this unique approach is likely to provide important insights into the pathways of tumour development and might ultimately lead to more effective individually targeted cancer-prevention strategies.

Genetic susceptibility to infectious disease: lessons from mouse models of leishmaniasis

Susceptibility to infectious disease is influenced by multiple host genes, most of which are low penetrance QTLs that are difficult to map in humans. Leishmaniasis is a well-studied infectious disease with a variety of symptoms and well-defined immunological features. Mouse models of this disease have revealed more than 20 QTLs as being susceptibility genes, studies of which have made important contributions to our understanding of the host response to infection. The functional effects of individual QTLs differ widely, indicating a networked regulation of these effects. Several of these QTLs probably also influence susceptibility to other infections, indicating that their characterization will contribute to our understanding of susceptibility to infectious disease in general.

Genetic composition of the recombinant congenic strains

For the study of biological phenomena influenced by multiple genes in mice, the Recombinant Congenic Strains (RCS) have been developed. An RCS series comprises approximately 20 homozygous strains, each of which contains on average 87.5% genes of a common background strain and 12.5% of a common donor strain. In an RCS series, non-linked genes involved in the control of a multigenic trait become distributed into different re-combinant congenic strains. In this way a multigenic trait is transformed into a series of single gene traits in which each gene can be studied individually. For the ability to use the strength of the recombinant congenic strains system to its full extent, a thorough genetic characterization is indispensable. We have typed the CcS/ Dem and OcB/Dem series for 611 and 550 markers, respectively. This results in a genetic characterization sufficient to detect most donor strain genes. In addition, we report the genetic characterization of the HcB/Dem and HcB(N4)/Dem series. Strains of the latter series contain on average 6.25% of the donor strain genome. Both series have been typed for 130 markers. All the typing data have been deposited in the Mouse Genome Database at The Jackson Laboratory.

LOH of PTPRJ occurs early in colorectal cancer and is associated with chromosomal loss of 18q12–21

Recently, the gene PTPRJ (protein tyrosine phosphatase receptor type J) was identified as the candidate gene for the mouse colon cancer susceptibility locus Scc1. Its human homologue PTPRJ is frequently deleted in several cancer types, including colorectal cancer. To elucidate the role of PTPRJ loss in different stages of colorectal cancer and in its pathways of progression, we expanded the previously published comparative genomic hybridization results with novel data on loss of heterozygosity (LOH) at the PTPRJ locus. We identified a strong association between the LOH of PTPRJ and the loss of chromosomal region 18q12–21 (P=0.009). This observation is specific for progressed colorectal adenomas, suggesting that an interaction between LOH of PTPRJ and loss of 18q12–21 may be involved in the development of a more progressed form of adenomas.

The recombinant congenic strains for analysis of multigenic traits: genetic composition.

The genetic control of susceptibility to many common diseases, including cancer, is multigenic both in humans and in animals. This genetic complexity has presented a major obstacle in mapping the relevant genes. As a consequence, most geneticists and molecular biologists presently focus on “single gene” diseases. To make the multigenic diseases accessible to genetic and molecular analysis, we developed a novel genetic tool, the recombinant congenic strains (RCS) in the mouse (4). The RC strains are produced by inbreeding of mice of the second backcross generation between two inbred strains, one of which serves as the “donor” and the other as the “background” strain. A series of RCS consists of approximately 20 strains, each carrying a different set of genes: approximately 12.5% genes from the common donor inbred strain, the remaining 87.5% from the common background inbred strain. As the set of donor strain genes in each RC strain is different, the nonlinked genes of the donor strain involved in the control of a multigenic trait, e.g., cancer susceptibility, become distributed into different RC strains where they can be analyzed one by one. Hence, the RCS system transforms a multigenic trait into a series of single gene traits, where each gene contributing to the multigenic control can be mapped and studied separately. Recently we demonstrated that the RCS system is indeed capable of resolving multigenic traits, which are hardly analyzable otherwise, by mapping four new colon tumor susceptibility loci (8; P. C. Groot, C. J. A. Moen, W. Dietrich, L. F. M. van Zutphen, E. S. Lander, and P. Demant, unpublished results). For successful application of the RCS system, extensive genetic characterization of the individual recombinant congenic strains is essential. In this paper we present detailed information about the genetic composition of three series of RC strains on the basis of typing of 120‐180 markers distributed along all autosomes. The data indicate that the relative representation of the donor strain genes in the RC strains does not deviate from the theoretical expectation, and that the RC strains achieved a very high degree of genetic homogeneity and for all practical purposes can be considered inbred strains. The density and distribution of markers reported here permits an effective mapping of unknown genes of donor strain origin at almost all autosomal locations. Much of this information has been obtained using the new class of genetic markers, the simple sequence repeat polymorphisms. The extensive application of these PCR‐typable markers greatly increases the possibilities of the effective use of the RC strains for mapping of genes involved in multigenic control. The data presented here can be used by investigators interested in functional effects of a specific region of the mouse genome to select RC strains that differ at this region from the background strain. This information is of interest to investigators working with mice, as well as those analyzing a specific region of the human genome and interested in the structure or function of the homologous part of the mouse genome.— Groot, P. C.; Moen, C. J. A.; Dietrich, W.; Stoye, J. P.; Lander, E. S.; Demant, P. The recombinant congenic strains for analysis of multigenic traits: genetic composition. FASEB J. 6: 2826‐2835; 1992.

The recombinant congenic strains—a novel genetic tool applied to the study of colon tumor development in the mouse

The development of tumors in mice is under multigenic control, but, in spite of considerable efforts, the identification of the genes involved has so far been unsuccessful, because of the insufficient resolution power of the available genetic tools. Therefore, a novel genetic tool, the RC (Recombinant Congenic) strains system, was designed. In this system, a series of RC strains is produced from two inbred strains, a “background” strain and a “donor” strain. Each RC strain contains a different small subset of genes from the donor strain and the majority of genes from the background strain. As a consequence, the individual genes of the donor strain which are involved in the genetic control of a multigenic trait, become separated into different RC strains, where they can be identified and studied individually. One of the RC strains series which we produced is made from the parental strains BALB/cHeA (background strain) and STS/A (donor strain). We describe the genetic composition of this BALB/cHeA-C-STS/A (CcS/Dem) series and show, using 45 genetic autosomal markers, that it does not deviate from the theoretical expectation. We studied the usefulness of the CcS/Dem RC strains for analysis of the genetics of colon tumor development. The two parental strains, BALB/cHeA and STS/A, are relatively resistant and highly susceptible, respectively, to the induction of colon tumors by 1,2-dimethylhydrazine (DMH). The individual RC strains differ widely in colon tumor development after DMH treatment; some are highly susceptible, while others are very resistant. This indicates that a limited number of genes with a major effect are responsible for the high susceptibility of the STS strain. Consequently, these genes can be mapped by further analysis of the susceptible RC strains. The differences between the RC strains were not limited to the number of tumors, but the RC strains differed also in size of the tumors and the relative susceptibility of the two sexes. Our data indicate that the number of tumors and the size of tumors are not controlled by the same genes. The genetics of these different aspects of colon tumorigenesis can also be studied by the RC strains. The DMH-treated mice of the parental strains and the RC strains also developed anal tumors and haemangiomas in varying numbers. The strain distribution pattern (SDP) of susceptibility for each of the three types of tumors induced by DMH is different, indicating that development of these tumors is under control of different, largely non-overlapping, sets of genes. Thus, with a single series of RC strains, genes involved in tumorigenesis in various organs and tissues can be studied separately. These results indicate that the novel genetic tool, the RC strain system, offers new possibilities for analysis of the multigenic control of tumor development.

Molecular Heterogeneity of H-2 Antigens

Since the discovery of the H-2Ld molecule (Lemonnier et al. 1975) we have demonstrated that several K and D region alleles produce more than one type of H-2 molecules. Two of four different molecules were distinguished in the products of different alleles. Some of these molecules are products of different genes (H-2D, H-2L), in other instances the evidence for distinct genes is not available. Some of the different molecules produced by the same region might be modified products of the same gene. In the instances where no information implicating different genes is available, we use a neutral terminology which does not presume a genetic difference: H-2K1d and H-2K2d, H-2D1k, H-2D2k, H-2D1dx, H-2D2dx, H-2L1d, H-2L2d, etc. Immunoprecipitation experiments with some anti-H-2L and anti-Qa-2 sera revealed proteins with the apparent molecular weight of 41,000. We designate these antigens provisionally Lq and Qx, respectively. The Lq protein is polymorphic and it is at least partly under the control of H-2L-linked genes since it is absent from BALB/c-H-2dm2 cells. Since we have never seen the 41,000 proteins in precipitates of H-2K or H-2D antigens, it appears that whatever the origin of these molecules, they reveal some features common to products of L and Qa region. The basic relationship of H-2 K, D, L antigens is revealed also by the shared antigenic specificities between these H-2 molecules which we demonstrate using anti-H-2.28 sera. In summary, our results show that the class I antigens in each haplotype represent a family of several distinct but antigenically related molecules. The specificities of the H-2.28 family are the strongest allotype common to different H-2 K, D, and L molecules. Recent direct demonstration of several different genes in the Dd region (Steinmetz et al. 1981) provides evidence for the genetic complexity of H-2 genes which may be underlying basis of the molecular heterogeneity of H-2 antigens discussed here.