Background: Cell lines are often poorly characterized from a genetic point of view, reducing their usefulness as tumor models. Our purpose was to assess the genetic background of eight commonly used human thyroid carcinoma models and to compare the findings with those reported for primary tumors of the gland.
Methods: We used chromosome banding analysis and comparative genomic hybridization to profile eight non-medullary thyroid carcinoma cell lines of papillary (TPC-1, FB2, K1 and B-CPAP), follicular (XTC-1) or anaplastic origin (8505C, C643 and HTH74). To assess the representativeness of the findings, we additionally performed a thorough review of cytogenetic (n = 125) and DNA copy number information (n = 270) available in the literature on clinical samples of thyroid carcinoma.
Results: The detailed characterization of chromosomal markers specific for each cell line revealed two cases of mistaken identities: FB2 was shown to derive from TPC-1 cells, whereas K1 cells have their origin in cell line GLAG-66. All cellular models displayed genomic aberrations of varying complexity, and recurrent gains at 5p, 5q, 8q, and 20q (6/7 cell lines) and losses at 8p, 13q, 18q, and Xp (4/7 cell lines) were seen. Importantly, the genomic profiles were compatible with those of the respective primary tumors, as seen in the meta-analysis of the existing literature data.
Conclusion: We provide the genomic background of seven independent thyroid carcinoma models representative of the clinical tumors of the corresponding histotypes, and highlight regions of recurrent aberrations that may guide future studies aimed at identifying target genes. Our findings further support the importance of routinely performing cytogenetic studies on cell lines, to detect cross-contamination mishaps such as those identified here.