Thyroid nodules are common in the adult human population. Many are benign but some are cancers with varying degrees of malignant potential. Clinical assessment cannot stratify individual thyroid nodules into definite benign, pre-malignant, or malignant categories. Pre-operative pathologic classification is imprecise because thyroid follicular hyperplasias, adenomas, and carcinomas have overlapping morphologic features. As a result, the majority of thyroid nodules resected at surgery are benign. We hypothesize that investigations into the pathogenesis of thyroid follicular neoplasms will elucidate their basic biology, improve diagnosis, identify reliable prognosticators, and reduce the need for thyroid surgeries. t(2;3)(q12;p25) has been identified in 8 of 11 human thyroid follicular carcinomas using cytogenetic and fluorescence in situ hybridization (FISH) analyses. t(2;3)(q12;p25) appears specific for follicular carcinoma in that it was absent from 12 papillary carcinomas, 21 follicular adenomas, and 12 multi-nodular hyperplasias lacked t(2;3)(q12;p25) analyzed by FISH. We hypothesize that t(2;3)(q12;p25) results in formation of a novel human oncogene that we propose to characterize and determine whether it is a marker identifying patients with malignant and pre-malignant follicular thyroid nodules. The t(2;3)(q12;p25) oncogene will be localized by physical mapping with YAC- and BAC-FISH of both breakpoints. Fusion oncogene cDNAs will be isolated with BACs and verified by: (1) their ability to hybridize to DNA spanning the translocation breakpoint(s), (2) identification of genomic DNA rearrangements and unique mRNA transcripts in follicular carcinoma tissue, and (3) direct nucleotide sequencing. In vitro systems will be established to characterize oncogene transforming activity, and the utility of the oncogene in diagnosis and classification of thyroid follicular neoplasms will be evaluated in histologic and cytologic pathology specimens.