Thyrotropin dependent and independent thyroid cell lines selected from FRTL-5 derived tumors grown in nude mice.

FRTL-5 cells were used to set up a thyroid tumor model system in C3H nu/nu mice. FRTL-5 tumors could be grown in nude mice provided serum TSH levels were elevated. Persistent TSH elevation was obtained by administration of Na131I, rendering the mice hypothyroid. After 4 weeks FRTL-5 cells were injected sc resulting in tumor growth within 2 weeks in eight out of eight mice. Although the tumors showed an apparently undifferentiated histology, lacking normal follicular structures, they were functional since the tumors were capable of concentrating [131]iodine, as demonstrated by nuclear imaging. From one of the tumors a new cell line was isolated (FRTL-5/T) that, like the parental FRTL-5 cell line, was TSH dependent for growth. In a control group of six euthyroid nude mice FRTL-5 tumor growth could not be obtained with one exception. After 3 months one animal developed a small tumor that grew rapidly thereafter. This tumor was easily transplantable in other euthyroid nude mice, showed an undifferentiated histology, and was nonfunctional, as it could not concentrate [131]iodine. From this tumor two cell lines were derived: one cultured in the presence of TSH (FRTL-5/TP) and one in the absence of TSH (FRTL-5/TA). Both cell lines were found to be TSH independent for growth. The cell lines were analyzed for TSH responsive functions and TSH receptor expression. Responsiveness to TSH in FRTL-5/T and the parental FRTL-5 cell line were similar for most thyroid specific functions tested. However, FRTL-5/T was less sensitive than FRTL-5 for TSH induced [3H]thymidine incorporation. Both cell lines had two classes of TSH binding sites with high and low affinity respectively, as determined by Scatchard analysis. FRTL-5/TP and FRTL-5/TA were both able to grow in TSH free medium and were nonresponsive to TSH in vitro, as tested for [3H]thymidine and [3H]uridine incorporation, iodine uptake, thyroglobulin iodination, and thyroglobulin secretion. This correlated with an approximately 100-fold decreased number of TSH binding sites compared to FRTL-5. The latter was caused by a complete absence of low affinity binding sites, whereas high affinity receptors were still detectable. The FRTL-5/TA cell line was the least differentiated one as thyroglobulin mRNA was detectable in only minute amounts and thyroid peroxidase expression could not be measured. These in vivo selected FRTL-5 cell lines offer a suitable model to investigate several aspects of TSH responsiveness, including signal transduction and postreceptor events, thyroid differentiation, and thyroid tumorigenesis.