Abstract
The study by Romitti et al. (6) published in this issue of Thyroid demonstrates increased expression of D3 and reduced D2 expression in papillary thyroid carcinoma (PTC), and to a lower extent also in follicular thyroid carcinoma (FTC). The highest D3 expression was demonstrated in tumors with a more aggressive clinical behavior and in tumors with a BRAF V600E mutation, suggesting that low local T3 levels represent a growth advantage for differentiated thyroid cancer (DTC).
These data are in line with previous studies in which increased D3 expression was shown in other types of cancer, such as basal cell carcinoma (BCC) of the skin (7) and hemangioma (8). Independent of deiodinase expression, hypothyroidism enhances tumor invasiveness and metastasis development in nude mice inoculated with tumor cells (9). Furthermore, TRβ1 has been shown to act as a potent suppressor of tumor invasiveness and metastasis in hepatocellular carcinoma and breast cancer cells (10) by suppressing the activation of extracellular signal-regulated kinase and phosphatidylinositol 3-kinase signaling pathways that are crucial for cell proliferation and invasiveness.
In general, tissue D3 expression is high during fetal development and decreases markedly after birth, whereas the opposite is true for tissue D2 expression (1). The low cellular T3 levels resulting from the high D3 and low D2 activities are thought to be essential for the growth of fetal tissues. The timely differentiation of tissue cells is induced by the increased cellular T3 levels resulting from the down-regulation of D3 and the up-regulation of D2 expression. It is therefore not surprising that cell dedifferentiation during carcinogenesis is associated with increased D3 and decreased D2 expression. The cause–effect relationship between D3 expression and tumor aggressiveness may be bidirectional. The magnitude of the D3 induction may depend on the degree of dedifferentiation of the tumor cells, but the consequently low T3 levels may also contribute importantly to an increased proliferation rate and invasiveness of the tumor. The increased expression of D3 during thyroid carcinogenesis suggests that D3 also plays a role during thyroid development. Indeed marked expression of D3 has been observed in the developing thyroid in Xenopus (11).
One of the most intriguing aspects of D3 is that it is encoded by a gene (DIO3) located in the imprinted DLK1-DIO3 region on human chromosome 14q32 (12). The two protein-coding genes in this locus, delta-like homolog 1 (DLK1) and DIO3, are largely expressed from the paternal allele. In addition to these protein-coding genes, the DLK1-DIO3 locus contains many genes coding for short RNAs (snoRNAs and miRNAs), the exact roles of which remain to be determined. The DLK1-DIO3 locus is activated in stem cells and also in cancer stem cells (13,14). DLK1 is a growth factor that inhibits Notch signaling (15). It is highly expressed in progenitor cells, and its down-regulation results in the differentiation of different cell types. It would be interesting to know if DLK1 expression is also increased like D3 in DTC.
In addition to the increased D3 and decreased D2 expression as observed by Romitti et al. (6), thyroid carcinogenesis is also associated in general with a loss of TH synthesis capacity, further lowering cellular T3 levels and potentially increasing tumor aggressiveness. Conversely, an inhibitory effect on tumor growth may be expected by increasing cellular T3 levels in DTC, as has been shown in BCC (7). It is tempting to speculate about the potential therapeutic options of a specific D3 inhibitor, which may increase local T3 levels in thyroid tumors and other D3-expressing tumors and thereby decrease their growth and invasiveness. Interestingly, patients with metastatic DTC are treated with supra-physiological doses of T4 to achieve thyrotropin (TSH) suppression (16), but it is unknown to what extent circulating TH is taken up by thyroid tumor cells. The thyroid hormone transporter MCT8 is highly expressed in the thyroid from a very early stage (17). In the normal thyroid, MCT8 appears to play an important role in TH secretion (18,19), but it may also be involved in TH uptake in thyroid tumor cells.
The level of D3 activity in PTC is relatively low, compared with the levels described, for example, in patients with hemangiomas (8). The very high D3 expression in these tumors in combination with a high tumor load may represent such a large T3-degrading capacity that this results in “consumptive hypothyroidism.” The level of D3 activity in PTC is similar to what has been described in skeletal muscle during critical illness (20). In skeletal muscle, even these relatively low levels of D3 activity may contribute to low local as well as systemic T3 levels due to the abundance of muscle tissue (20,21). Although a systemic effect of the D3 activity in local PTC is highly unlikely, it would be interesting to study if patients with widely metastasized DTC have altered serum iodothyronine levels and if they require increased T4 doses to achieve adequate levels of T4 and T3 to suppress TSH.
The study by Romitti et al. (6) demonstrate the highest D3 activities in patients with more advanced or more aggressive PTC, suggesting that D3 may be a marker of dedifferentiation in these tumors. In this respect, the highest levels of D3 would be expected in patients with anaplastic thyroid carcinoma. Surprisingly, no D3 staining could be detected in these carcinomas. No D3 expression could be detected in medullary thyroid carcinoma either, probably reflecting the different origin of medullary thyroid cancer cells.
Since D3 is barely expressed in normal thyroid follicular cells, it is interesting to speculate whether D3 could be used as a diagnostic tool when fine-needle aspiration shows inconclusive results. This would be most useful for the diagnosis of FTC. Unfortunately, D3 expression in FTC was much lower than in PTC, and D3 was not completely absent in normal thyroid tissue. Nevertheless, the field of novel molecular markers in the diagnosis of thyroid cancer by fine-needle aspiration is rapidly evolving (22), and perhaps D3 will be one of several markers that may help to discriminate between benign and malignant thyroid disease. Alternatively, D3 could be used as a marker of aggressiveness in patients with already diagnosed PTC after thyroid surgery.
How D3 is induced in DTC, and why D3 levels are much higher in PTC than in FTC, is yet unclear. The observation by Romitti et al. (6) that D3 activity is highest in PTCs with a BRAF V600E mutation, suggests involvement of the mitogen activated protein kinase (MAPK) pathway. This would imply that the mechanism of D3 induction is different in different tumors since it has been shown that D3 induction in BCC is mediated via constitutive activation of the Sonic hedgehog pathway (7). In other pathophysiological conditions, such as heart failure and nonthyroidal illness, D3 has been shown to be induced via hypoxia inducible factor-1α (HIF-1α) and/or interleukin-6 (23 –25). Since the HIF-1α signaling pathway is induced in many cancers, it may also be involved in the up-regulation of D3 expression in different tumors.
In conclusion, more and more data on the relation between TH action and cancer are becoming available. The current study by Romitti et al. (6), suggesting that local hypothyroidism may contribute to growth and invasiveness of differentiated thyroid tumors, is a valuable addition to the current evidence. Nevertheless, the consequences of this local hypothyroidism for thyroid tumor growth and invasiveness remain to be demonstrated, as well as the mechanism for how D3 is induced.