Retinoic acid receptor 13 (RAR18), which codes for a nuclear receptor for retinoic acid, is localized in a chromosomal region frequently deleted in lung cancer cells.The gene is expressed in normal lung tissue and in the majority of the cell lines derived from lung tumors but not in most of the lines derived from lung tumors with epidermoid characteristics. To study the possible role of RAR.8 in growth control of epidermoid lung tumor-derived cells, transfectants expressing RARI3 were generated from nonexpressing epidermoid tumorderived cell lines. Four clones were derived from line CALU-1, three of which showed a 20-60% increase in doubling time in the presence of retinoic acid. Parental and control-transfected cells were unaffected or slightly stimulated. All four clones expressing RARj3 were less tumorigenic in nude mice than were the untransfected or control-transfected cells, with about a 50% incidence of take vs. 95%. When tumors did develop from RAR1-positive cells, they showed a reduced rate of growth, an increased latency, and, in six of seven tumors tested, a much reduced level of RARfi expression. Transfectants derived from a second tumor line, H157, also showed a markedly reduced incidence of take in nude mice. Together with the known effects of retinoic acid on differentiation and carcinogenesis, our results support the hypothesis that RAR.3 functions as a tumor suppressor gene in epidermoid lung tumorigenesis.

Genome-wide association (GWA) studies offer a powerful unbiased method for the identification of multiple susceptibility genes for complex diseases. Here we report the results of a GWA study for Crohn's disease (CD) using family trios from the Quebec Founder Population (QFP). Haplotype-based association analyses identified multiple regions associated with the disease that met the criteria for genome-wide significance, with many containing a gene whose function appears relevant to CD. A proportion of these were replicated in two independent German Caucasian samples, including the established CD loci NOD2 and IBD5. The recently described IL23R locus was also identified and replicated. For this region, multiple individuals with all major haplotypes in the QFP were sequenced and extensive fine mapping performed to identify risk and protective alleles. Several additional loci, including a region on 3p21 containing several plausible candidate genes, a region near JAKMIP1 on 4p16.1, and two larger regions on chromosome 17 were replicated. Together with previously published loci, the spectrum of CD genes identified to date involves biochemical networks that affect epithelial defense mechanisms, innate and adaptive immune response, and the repair or remodeling of tissue.haplotype ͉ complex disease ͉ IL23R C rohn's disease (CD) is a chronic inflammatory bowel disease characterized by transmural inflammatory lesions that can affect the entire gastrointestinal tract (1). The lifetime prevalence is 0.5-1% in Caucasian populations (2) and reflects the combined effects of genetic predisposition and environmental factors (3). Genetic linkage and candidate gene approaches (4-14) have contributed to the elucidation of loci influencing genetic susceptibility to CD. More recently, genome-wide association (GWA) studies (15-21) have provided further insight into the molecular pathogenesis of the disease. The top candidate genes or loci that consistently replicate include NOD2, IL23R, ATG16L1, the IBD5 region on chromosome 5q31, and a region on 5p13.1 near the PTGER4 gene. The nature of these genes suggests that the major genetic risk factors for CD are involved in the innate immune response and destruction of intracellular bacteria.It is now clear that GWA studies provide a powerful and robust new tool for the identification of the multiple susceptibility alleles involved in complex diseases. Importantly, these types of studies have the ability to identify genes that impart only moderate increases in risk (21,22). However, most studies performed to date have identified only a few top signals, and the validation of true association among signals with lower statistical significance remains a challenge. In addition, most of the GWA studies to date have been performed by using general populations, for which very large sample sizes are required for success. They have also largely relied on single-marker analysis, with genome-wide haplotype-based association analyses receiving little attention.In early 2004, we conducted a GWA study for CD...

No abstract