FINE MAPPING AND CANDIDATE GENE ANALYSES OF MURINE LUNG TUMOR SUSCEPTIBILITY GENES DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Min Wang, B.S. ***** The Ohio State University 2003 Dissertation Committee: Approved by Professor Ming You, Advisor Advisor Professor Christoph Plass, Co-advisor Professor Gary Stoner Co-advisor Molecular Virology, Immunology and Professor Yian Wang Medical Genetics Graduate Program ABSTRACT Lung cancer is the leading cause of cancer death in men and women in the United States. While the exposure to tobacco smoking and other environmental carcinogens represents the main risk factors, it has been clear that inherited components may be also involved in the development of lung cancer. Segregation analysis has revealed that susceptibility of the human population to different forms of lung cancer follows a pattern of autosomal dominant Mendelian inheritance. However, due to the genetic heterogeneity, substantial variance in exposure to environmental risk factors, poor prognosis and other factors, identification of human lung cancer susceptibility genes by direct linkage analysis is difficult. Mouse inbred strains show widely different susceptibilities to both spontaneously occurring and chemically induced lung tumorigenesis and become a valuable model to study lung cancer genetics. To date, linkage analyses using various mouse crossing panels have uncovered more than 20 putative lung tumor susceptibility /resistance loci on mouse chromosomes. Considering the highly homologous relationship between mouse and human genomes, identifications of these putative loci may ultimately lead to the discovery of human lung cancer susceptibility genes. A quantitative trait locus (QTL), Pulmonary adenoma susceptibility 1 (Pas1), was previously mapped to the distal region of mouse chromosome 6 using genetic linkage ii analyses on crosses between susceptible A/J and resistant C3H/He or C57BL/6J strains. This locus accounts for approximately 40-60% variance in lung tumor multiplicity induced by chemical carcinogens between the A/J and C3H/He or C57BL/6J mice. In addition to predisposing lung tumor multiplicity, the Pas1 locus also affects mouse lung tumor sizes. Although it is the major QTL in predisposition to mouse lung tumor susceptibility, the Pas1 gene(s) has not been identified. In the present studies, we have fine-mapped the Pas1 locus using two independent strategies. We have utilized a newly developed F11 generation of Advanced Intercross Line (AIL) mouse population to fine map Pas1, Pas2 (on chromosome 17) and Pas3 (on chromosome 19) QTLs. By selectively genotyping 30% of the population, we have confirmed the Pas1 QTL and refined it into an interval of approximately 1.0-cM (~1.3 Mb) in the vicinity of the Kras2 gene. The Pas2 QTL was detected by both ANOVA and regression analysis but not by Mapmaker software. An interaction between the Pas1 and Pas2 QTLs was also revealed. However, the Pas3 QTL has not been confirmed in this study. Congenic strategy was also utilized to fine map the Pas1 QTL. Like in AIL project, we started with two parental strains: lung tumor susceptible A/J strain and lung tumor resistant C57BL/6J strain. After seven generations of backcrossing, N7 congenic mice that carried ~36Mb of the A/J Pas1 QTL region were further crossed to the C57BL/6J mice to generate subcongenic strains of mice (N8), which contain various iii Pas1 QTL regions. N9 mice, generated by mating each selected N8 male mouse with three C57BL/6J female mice, were treated with an initiating dosage of carcinogens and were allowed to form lung tumors. A set of new polymorphic markers has been developed to assist this fine-structure mapping. Combining results from the AIL project and the congenic project, we have refined the Pas1 QTL into a less than 1-Mb minimum candidate region based on the Celera/or public mouse genome maps encompassed by the markers D6Osu6 and D6Osu11. Based on the fine-mapping results, the Pas1 locus was then sufficiently fine- mapped that candidate gene screening for the Pas1 locus could be performed. After initial screening, six genes located in the minimum Pas1 candidate region were selected for further examinations. The Lrmp/Jaw1 gene bears amino acid polymorphisms tightly co- segregating with strain Pas1 alleles. The RIKEN Ak016641 (re-named as Pas1c1 for Pas1 candidate 1) gene, encoding an intermediate filament tail domain-containing protein, produces alternatively spliced transcripts in inbred strains of mice and also Mus Spretus. Its mRNA splicing pattern tightly co-segregates with Pas1 alleles. Another novel gene, Pas1c2, identified in our study, was found bearing an amino-acid changing nucleotide polymorphism between lung tumor susceptible and resistant strains, which is also tightly correlated with strain Pas1 status. Thus, our results support these three genes as strong candidates for the Pas1 QTL, and they are being further tested by in vitro and in vivo functional analyses. We also examined the other three genes (Eca39, RIKEN iv Ak015530 and mHoj-1). Neither functional polymorphism nor expression difference has been found for Eca39 and mHoj-1 genes between lung tumor susceptible and resistant strains. The Ak015530 gene carries an amino-acid polymorphism but this polymorphism does not co-segregate with mouse lung tumor susceptibility. Thus, these three genes are less likely candidates for the Pas1 locus. Genetic studies have mapped a lung tumor resistance locus designated as Par2 to mouse chromosome 18. Par2 accounts for approximately 40% of the difference in lung tumor multiplicity between the susceptible A/J mice and the relative resistant BALB/c mice. We have fine mapped the Par2 locus using congenic mice that were constructed by placing an approximately 28-cM Par2 QTL region from the A/J mice onto the genetic background of BALB/c mice. Subcongenic strains of mice (N8) containing various Par2 QTL regions were generated and genotyped. Like in Pas1 congenic project, the N9 mice were generated by mating each selected N8 male mouse with three BALB/c female mice and were treated with urethane to induce lung tumor formation. Consequently, by analyzing the lung tumor multiplicity and genotypes of each subcongenic strain, the Par2 locus was narrowed to an approximately 6.3-Mb region flanked by the marker D18Mit103 and the marker D18Mit162. A mouse/human comparative map was constructed for the Par2 candidate region based on Celera genomic information. A high homology between the mouse Par2 candidate region and the human syntenic region on chromosome 18q21 has been observed. Based on recent mouse genome maps, there are v totally 79 putative genes residing in the candidate region. A subset of these genes was found to exhibit differential gene expression in lungs between the A/J and BALB/c mice when assayed by real-time RT-PCR. Four genes are possible Par2 candidates based on this assay, including Myo5b, Smad7, Mapk4, and GABA-A receptor-like gene mCG58197. More interestingly, sequencing analyses for this region found that the Rad30b gene, encoding the DNA–dependent polymerase iota (Polι), carries 25 nucleotide polymorphisms in its coding region between A/J and BALB/c inbred strains, causing ten amino-acid alterations. Functional analyses on the above five genes are needed to clarify their Par2 candidacy. In addition, 54 putative genes exhibited various types of SNPs when comparing genomic sequences from A/J, 129X1/svJ, 129S1/svImJ, DBA/2J, and C57BL/6J mice. Screening of these SNPs in the future may help us find polymorphic markers and identify new Par2 candidate genes. vi Dedicated to my family, my mentors and all my friends vii ACKNOWLEDGMENTS I wish to thank very much my advisor, Dr. Ming You, for his intellectual support, encouragement, and enthusiasm that made this thesis possible, and for his patience in correcting both my stylistic and scientific errors. I wish to thank my other advisory committee members, Dr. Christoph Plass, Dr. Gary Stoner and Dr. Yian Wang, for their continual devotion of precious time and energy to my Ph.D graduate training. I would like to thank our collaborators: Dr. Alvin M. Malkinson from University of Colorado, Dr. Fuad Iraqi from International Livestock Research Institute in Kenya and Theodora R. Devereux from National Institute of Environmental Health Science, for their intellectual and material supports to our projects. I am grateful to Dr. Zhongqiu Zhang, Dr. William J. Lemon and Dr. Futamura Manabu for their critical contribution to the present projects and intellectual discussion during my research. I appreciate the opportunity of working in such a wonderful laboratory. I thank my colleagues in this laboratory for all the helps I have received from them during my research. Finally, I cannot thank enough to my family for their both mental and physical support during my Ph.D. training period. viii VITA Dec.12, 1975………………………..Born, Hubei Province, P. R. China 1997…………………………………BS in Biology, South China Normal University, P. R. China 1997-1998……………….………….MS student, Dept. of Cancer Biochemistry, Guangzhou Medical College, P. R. China 1998-1999…………………………..MS student, Dept. of Pathology, Medical College of Ohio, Toledo, Ohio 1999-2000…………………………..Ph.D. fellowship, Molecular Basis of Disease Program, Medical College of Ohio, Toledo, Ohio 2000-present………………………...Ph.D. Graduate Research Assistant, Molecular Virology, Immunology and Medical Genetics Program, The Ohio State University, Ohio ix PUBLICATIONS Research Publication 1. Fine mapping of Pulmonary adenoma susceptibility1 (Pas1) gene using congenic mice. (Temporary title and in preparation) 2. Identification of Pulmonary adenoma resistance 2 (Par2) gene(s) on mouse chromosome 18. (Temporary title and in preparation) 3. Wang M, Lemon WJ, Liu GJ, Wang Y, Teale AJ, Iraqi F, Malkinson AM, and You M. Fine mapping of the pulmonary adenoma susceptibility gene 1 (Pas1) locus using advanced intercross lines. Cancer Research 2003 Jun 15, 63 (12) 4.