GPCR5A knockout mouse: A new model for lung carcinogenesis.

ED06-02 Lung carcinogenesis is a multistep process involving progressive genetic and epigenetic changes. Tobacco smoke is the major etiological agent, causing more than 85% of lung cancers. Non-small cell lung cancers (NSCLC) are thought to develop via morphologically distinct premalignant lesions by a multistage process. The initiation and progression of these lesions involve both activation of oncogenes and inactivation of tumor-suppressor genes. Mouse models for human cancer, especially those that reproduce genetic changes that underlay the human disease have proven to be valuable tools for understanding the basic tumor biology as well as for the development and validation of new approaches to cancer prevention and therapy. Most mouse strains do not develop or have a low incidence of spontaneous lung cancer, however, mice of the strain A/J are unique in that the majority (90%) develop lung adenoma, which may progress infrequently to adenocarcinoma as the mice reach 18 month of age. Such spontaneous lung tumors uniformly possess a mutant K-ras oncogene. When the mice are exposed to tobacco smoke carcinogens like 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) lung tumorigenesis is enhanced through promotion of the spontaneously occurring K-ras mutations. Retinoids (e.g., all-trans-retinoic acid, ATRA), play important physiological roles in lung embryonal development, maintenance of mucociliary epithelial differentiation, and possibly also as endogenous inhibitors of lung carcinogenesis. Retinoids possess both chemopreventive and therapeutic effects in animal models of lung carcinogenesis. These effects are mediated by activation of nuclear retinoid receptors, which are ligand-dependent enhancers of target gene transcription. We were interested in the identification of retinoic acid target genes because we hypothesized that some of them might mediate the effects of retinoic acid on cell proliferation, survival and chemopreventive potential and might be new targets for intervention. Using differential display, we identified a novel gene, designated GPRC5A (synonyms: RAIG, RAI3, GPCR5A), which was induced by retinoic acid in human head and neck and non-small cell lung carcinoma (NSCLC) cell lines. The gene encodes a protein with 7 transmembrane domains, which appears to be an orphan G protein coupled receptor. We also cloned the mouse homolog (Gpcr5a), which is similar to the human gene. Both are expressed preferentially in the lung. Several NSCLC cell lines showed reduced expression that could be restored by retinoic acid. Furthermore, an almost identical retinoic acid response element to which retinoid receptors bind was identified in the promoter region of both the mouse and human genes. The GPCR5A mRNA level was reduced in more than 60% of 18 paired human adjacent normal lung tissue and non-small cell lung cancers (both adenocarcinomas and squamous cell carcinomas). Its overexpression in several cell lines suppressed anchorage-independent colony formation. Thus, it seemed that GPCR5A expression is inversely related to lung cancer and to suppress the transformed phenotype. To better understand the function of the Gprc5a gene and protein, we deleted it (knockout) in mouse using gene-targeting technique. We replaced the Gpcr5a gene with the reporter bacterial beta-galactosidase (Lac Z) gene in mouse embryonal stem cells (ES) and used them to generate mice (strain B6;129F1) deficient in Gpcr5a. The heterozygous or homozygous mice showed no developmental aberrations and were as fertile and gained weight as the wild type littermates. The mice were interbred and followed over a period of up to 2 years for signs of abnormalities. A subset of mice was injected at age 2 months with tobacco carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and sacrificed 17 months later. A pathologist analyzed the tissues after HE (+/-); and (-/-) mice was 2% (1/49), 15% (14/93), and 63%(28/44), respectively. A microscopic analysis of histological sections prepared from formalin fixed paraffin-embedded lungs revealed that 75% of the tumors in the (-/-) mice were (premalignant) lung adenomas and a few were adenocarcinomas. The few tumors in heterozygotes and wild type mice were all adenomas. Histological sections of some adenomas and adenocarcinomas were analyzed by immunohistochemical techniques for the expression of the differentiation markers CC10 for Clara cells or type I cell, surfactant proteins A, B, and C (SPA, SPB, and pro-SPC) for type II cells. The analysis revealed that most adenomas expressed the type II cell markers and none expressed the Clara cell marker. The NNK treated mice were killed at 17 months after NNK injection and the presence of lung tumors was assessed. 15/15 (100%) of NNK-treated Gpcr5a(-/-) mice developed lung tumors, compared to 8/14 (57%) of Gpcr5a(-/-) mice without NNK treatment (p = 0.0063). None of the wild-type littermates, whether treated with NNK (0/15) or not (0/20), had developed lung tumors in keeping with the expected carcinogen resistance of their parental strains. No Ras mutations were found in any of the tumors tested. Thus, the carcinomas found in Gpcr5a mice appear to develop via a Ras mutation-independent carcinogenesis process. These results support the conclusion that Gprc5a functions as a tumor suppressor gene in the lung and as an enhancer of NNK carcinogenesis. This novel animal model of lung carcinogenesis might be useful for improving the understanding of lung cancer development and for testing new chemopreventive and therapeutic agents.