Abstract
A recent report of Schweppe et al. (1) showed that many of the available thyroid cancer cell lines are genetically identical to cell lines of other origins, suggesting that they are not of thyroid lineage. Specific cell lines are useful to study the biological and molecular characteristics of the corresponding tissue cells. This is especially true for cancer, as stable cell lines are suitable for the study of tumor type–specific oncogenes, for growth factor signaling, and to test the efficacy of potential therapeutic agents. The use of cells in culture has increased as public opinion has placed constraints on the use of animal models. However, recent and past analyses on the nature of cell lines indicate that many are of a different origin and even species from that indicated, varying even with the laboratory where they are cultured (2,3). The mistaken identity of these cells invalidates the conclusions of experimental results, further complicating already existing difficulties in translating data obtained from tissue culture work to in vivo situations.
One possible cause of misidentification is cross-contamination between cell lines cultured in the same laboratory. The risk increases with the number of cell lines used in the same experiment. Besides human mistakes, we have to take into account the known genetic instability of cell lines, especially cancer cell lines submitted to long-term culture. This is particularly true for cell lines widely used for decades. As is the case of cellular heterogeneity within a single tumor, clonotypic heterogeneity in long-term cultured cell lines can be explained by the cancer stem cell hypothesis as well as the clonal evolution model (4,5).
Thyroid research is conducted by a large, worldwide, but cohesive community. This is facilitated by the scientific thyroid associations such as the European Thyroid Association, the American Thyroid Association, the Asian and Oceania Thyroid Association, the Latin American Thyroid Society, and the International Thyroid Congress, which provide research groups with knowledge and the opportunity to exchange research material. We, as well as many other investigators of this community, have obtained thyroid cell lines from colleagues instead of from a more controlled and certificated cell banks. In turn, we have shared the same cells with other laboratories. Thus, we obtained and gave to other laboratories NPA, ARO, TPC-1, FRTL-5 along with other cell lines. The existence of clonotypic heterogeneity was already known for some of these cells. Even in the same institution, we had to rename the FRTL-5 cells that retain a close behavior to the very early cells established in the 1980s, to distinguish them from more recent FRTL-5 subclones with less thyrotropin dependence that must be kept separately (6). Similarly, we have in our laboratory a subclone of ARO cells that is much more resistant to drugs and to pro-apoptotic stimuli in general than the homonymous clone used in the same institute (7,8). To date a great number of studies can be found in the literature in which NPA, ARO, or KAT cells were used. Because of the concerns noted here, it is apparent that a certified origin of the cells that are used for thyroid and other research is mandatory to provide new information and permit the most optimal interpretation of results. Not all data obtained in cells with ambiguous origin are useless, however. For example, cell lines such as COS, CHO, and NIH-3T3 are used to study diverse biological phenomena, from signaling to transcription, and from cell cycle to apoptosis, even though they are not considered as models of their original source, kidney, ovary, and fibroblasts, respectively. The efficacy of the therapeutic agents can be demonstrated in cells independently of their origin, provided the interpretation of the data is appropriate and confined to the observed molecular interactions.
With this letter, we wish to warn the international scientific community, and in particular editors and referees, that reports based on such ambiguous thyroid cellular models are being published in important and prestigious journals. Therefore, more attention should be paid at the design stage of experiments, if at all possible avoiding the use of cells whose origin is doubtful, and careful review and dialog in the publication phase so that critical discussion will be optimally informative when results with potentially questionable cells are obtained.