Peter Demant, MD, PhD

The main interest of the group is the study of the genes in host genome, which control the development of cancer. We aim to map and subsequently clone such genes, understand the way they interact with different molecular pathways during tumor development and progression, and determine what their relationship is to processes of regulation of cellular homeostasis at different developmental stages. The first evidence for the role of genetic factors in cancer has been obtained in familial cancer syndromes, which are caused by germ-line mutations in tumor suppressor or mismatch repair genes like APC, Rb, NF2, MSH2. However, these syndromes account for a small part of all cancers. The largest proportion of cancers is of non-familial, sporadic, type. Nevertheless, also in these apparently non-inherited cancers the host genetic factors play an important role (e.g., Lichtenstein et al., New Eng J Med 343:78, 2000). Moreover, especially in the breast cancer there is extended evidence that many, and probably most, cases of the apparently non-familial breast cancer appear in the genetically predisposed individuals (Peto and Mack, Nat Genet 26:411-414, 2000; Pharoah et al., Nat Genet 31:33-36, 2002). The largest part of this predisposition is caused by presently unknown genes. In contrast to the germ-line mutations of oncogenes or tumor suppressor and mismatch repair genes, the genes affecting development and progression of sporadic cancers are multiple and have smaller effects. Therefore, they are difficult to detect in classical family and population studies, and they have to be analyzed in experimental animals, which offer the advantage of defined genetic constitution, possibility of producing informative crosses, and standardized tumor-induction procedures (Demant, Nat Rev Genet 4:721-734, 2003). We used recombinant congenic strains of mice to map and identify lung and colon cancer susceptibility genes in mice and subsequently to search for their homologues in humans.

Project 1. Genetics of Colon Cancer Susceptibility

The cloning of cancer susceptibility genes becomes a subject of major interest for cancer geneticists. The first reported success in this respect is the Scc1 (Susceptibility to colon cancer 1) gene, identified as the Ptprj (Protein tyrosine phosphatase receptor-type J) (Ruivenkamp et al., 2002). Ptprj has one intracytoplasmatic enzyme domain, a transmembrane part and an extracellular chain consisting of fibronectin III-like domains. The Ptprj products of the susceptible and resistant strains differ at several amino acids in their fibronectin type III domains. Most of these amino-acid substitutions reside in the extracellular portion of the molecule. Similarly, the sequence of the coding regions of the homologous human gene (PTPRJ), which has been established in a large number of human DNAs, revealed several amino-acid replacements in extracellular regions. Sequence alignment, secondary structure prediction and homology modeling predict that the substitutions in the mouse as well as in humans occur in the exposed regions of the molecule that are available for interactions with ligands or other proteins, and thus could affect the signaling process. Presently, a Ptprj knockout mouse has been prepared and is being tested for its pattern of colon tumor susceptibility.

A large proportion of colon, breast and lung cancers exhibits loss of heterozygosity (LOH) at the short PTPRJ-bearing segment of chromosome 11p11 with their minimal shared segment containing PTPRJ In order to investigate the specific pathway and tumor progression stage, at which this gene may operate, we examined the possible correlation of the LOH in colon cancers with other somatic alterations. This analysis revealed a correlation between the LOH at PTPRJ and the deletion of the chromosomal segment 18q12-21 (Ruivenkamp et al., Oncogene 22:3472-3474, 2003) in progressed adenomas, indicating that haplodeficiency of PTPRJ is an early event in colon carcinogenesis and that its loss, in possible synergy with a deficiency of another gene on chromosome 18q, supports the transition of early adenomas into a more progressed stage. A significant definition of the susceptibility genotype requires definition of a large number of susceptibility genes. The search for novel colon cancer susceptibility genes revealed five novel Scc loci, Scc10 – Scc14 (Ruivenkamp et al., Oncogene 22:3472-3474, 2003).

Project 2. Lung Cancer Susceptibility

Lung cancer is the leading cause of cancer deaths worldwide. It exhibits familial clustering both in smokers and non-smokers, indicating the involvement of presently unknown genes. However, no such genes have been identified yet in humans. Using the mice as a model for study of genetics of lung cancer, the comparison of a number of mouse inbred strains revealed large differences in susceptibility to both spontaneous and chemically induced lung tumors and resulted in the mapping of several susceptibility loci. We have shown that the total number of Sluc genes (Susceptibility to lung cancer) probably exceeds 60 (Tripodis et al., 2001). Thirty of them have been mapped in the crosses of OcB/Dem RC strains, in which the genomes of the strain O20 and B10.O20 are segregated. OcB/Dem RCS have each only a random ~12.5% subset of genes derived from the resistant strain B10.O20, and the rest from the strain O20. In the past period, we used these strains to study the genetic control of lung cancer phenotype, because their respective parents, O20 and B10.O20, produce lung tumors with different biological properties. The microscopic appearance of cancer has been the major predictor of its clinical behavior. In order to be able to study the genes affecting lung tumor progression and correlate them with gene expression patterns, we used a new method for qualitative assessment of mouse lung tumors that provides a numerical description of the 3-D shape of a tumor. Eight new lung tumor shape determining (Ltsd) loci have been detected that influence the overall three dimensional (3D) shape of a tumor. We established the specific effects of these loci on presence of focal or regional regions of tumor progression. These studies lay the basis for a more detailed study of the mechanisms of genetic control of tumor progression (Tripodis and Demant, Cancer Res 63:125-131, 2003). These findings open way for expression profiling studies to establish the correlation between the host genotype, tumor phenotype, and tumor gene expression pattern.