Abstract
Thyroid-stimulating hormone (TSH) has long been recognized as the major proliferative and functional stimulus for thyroid follicular cells. TSH receptor (TSHR) engagement stimulates the production of cyclic AMP and the subsequent activation of downstream effector molecules, including protein kinase A, S6K1, and Rap1, whereas the role of the RAS and phosphatidylinositol-3-kinase signaling cascades downstream of TSHR is still controversial. Despite the abundance of candidates, it is still unclear which of these pathways represent(s) the key mitogenic output of TSH-initiated signaling. We have used an in vivo model of goitrogenesis to dissect the contribution of these pathways to TSH-induced thyrocyte proliferation and thyroid hyperplasia. We show that the in vivo proliferative response to chronic TSHR stimulation relies heavily on the activation of the mTOR/S6K1 axis, and that mTOR inhibition during goitrogenic stimulation abrogates the hyperplastic but not the hypertrophic thyrocyte responses to TSH, thus functionally uncoupling these two processes. Strikingly, goitrogenesis was not associated with an increase in AKT phosphorylation levels, underlining the existence of an AKT-independent pathway leading to mTOR activation upon TSH stimulation. [Cancer Res 2007;67(17):8002–6]
- Thyroid
- TSH
- Akt
- mTOR
- S6K1
- Goiter
Introduction
Thyroid function and proliferation are primarily induced by the thyroid-stimulating hormone (TSH). TSH is secreted by the pituitary and its levels are regulated by the thyroid hormones through a negative feedback mechanism. TSH binding to its receptor (TSHR) activates the Gαs/adenylate cyclase/cyclic AMP (cAMP) cascade, leading to both the induction of cell cycle progression and the expression of differentiation markers ( 1). Identification of the crucial nodes of the signaling cascades involved in thyrocyte proliferation is essential to understand the pathogenetic mechanisms of thyroid proliferative disorders, from simple goiter to thyroid carcinoma. Most data on the mitogenic pathways activated by TSH have been obtained in cell cultures from rat, dog, and human thyroids. In addition to some common themes, however, a number of contrasting results have been obtained, so that several open questions still remain ( 2).
The notion that TSH mitogenic signals are funneled through the cAMP pathway is supported by the fact that cAMP enhancers such as forskolin and cholera toxin reproduce the effects of TSH on cell proliferation in dog primary thyrocytes ( 3), and by mouse models based on the activation of the TSHR/cAMP cascade ( 4), suggesting that the activity of protein kinase A (PKA) is absolutely required for the mitogenic stimulation by TSH. Likely candidates to transduce these mitogenic signals include the cAMP-responsive element binding protein (CREB; ref. 5), p70S6K ( 6), and the small GTPase Rap1 ( 7). However, microinjection of Rap1 failed to induce proliferation of dog thyrocytes, showing that Rap1 alone is not sufficient to trigger thyroid cell growth ( 8).
The role of phosphatidylinositol-3-kinase (PI3K) in the TSH-induced mitogenic cascade is still unclear: whereas rat thyroid cells in culture seem to require PI3K activation downstream of TSH stimulation for proliferation ( 2), cAMP and PKA cannot activate PI3K or AKT in dog primary thyrocytes, a system in which the PI3K/AKT pathway does not seem to be involved in TSH-mediated thyrocyte proliferation ( 9). We have recently shown that constitutive activation of the PI3K/Akt cascade conveys a major proliferative signal in vivo in thyroid follicular cells ( 10). However, the relationship between this cascade and the signaling events downstream of TSHR activation still needs to be defined.
To elucidate the in vivo pathways that are activated by TSH during the induction of thyroid proliferation, we have used a well-established mouse model of chronic TSH stimulation. We provide strong evidence that the major TSH-generated mitogenic stimulus is conveyed through the mTOR kinase, but does not require AKT activation.
Materials and Methods
Animals and treatments. 129Sv mice were bred in the Fox Chase Cancer Center Laboratory Animal Facility. Hypothyroidism was induced at 4 to 6 weeks of age by the administration of 0.5% sodium perchlorate and 0.05% methimazole (both from Sigma), given in the drinking water for 5 days or 2 weeks. RAD001 [Everolimus; kindly provided by Dr. Heidi Lane (Novartis, Basel, Switzerland)], was administered daily by oral gavage at a dose of 10 mg/kg body weight. At the end of experiments, mice were euthanized and weighed. The thyroid was dissected and weighed; one lobe was fixed for pathologic analysis, and the other was frozen in liquid nitrogen for protein extraction. Experiments were repeated thrice.
Primary thyrocyte culture. Thyroids were dissected from 8-week-old mice and collected in 1 mL of Ham's F12 with 100 units/mL of type I collagenase (Sigma) and 1 unit/mL of dispase (Roche). Enzymatic digestion was carried out for 90 min at 37°C. After digestion, isolated follicles were washed thrice with culture medium (Ham's F12 containing 40% Nu-Serum IV; Collaborative Biomedical), seeded in a 12-well plate, and grown for 4 days. At day 4, the cells were switched to 4% Nu-Serum IV for 2 days. At day 6, cells were pretreated with vehicle or 20 μmol/L of RAD001 for 20 min and then stimulated with 2 mU/mL of TSH (Sigma) for 20 min, washed and lysed for Western blot analysis.
Immunohistochemical analysis. The following antibodies were used: rabbit polyclonal against Ki-67 (Vector Laboratories), rabbit monoclonal (immunohistochemistry-specific) against pAKT (Ser473) and pS6 (Ser235/236; Cell Signaling). Tissues were fixed, embedded in paraffin, and sectioned at 6 μm. Sections were subjected to antigen retrieval in 0.1 mol/L of sodium citrate and counterstained with hematoxylin.
Western blot. Thyroids were washed with PBS and homogenized on ice in cell extraction buffer (Invitrogen) supplemented with Complete proteinase inhibitor tablet (Roche). After centrifugation and protein determination, lysates (20 μg) were analyzed by SDS-PAGE. Western blot analysis was carried out with the following antibodies, all from Cell Signaling: AKT, pAKT (Ser473), pAKT (Thr308), S6K1, pS6K1 (Thr389), pS6K1 (Thr421/Ser424), S6, pS6 (Ser235/236), CREB, pCREB (Ser133), ERK1/2, and pERK1/2(Thr202/Tyr204).
Proliferation analysis. Ki-67–stained thyroid sections were photographed at 400× magnification and analyzed using the ImageJ software. Between 1,500 and 3,000 cells were analyzed per slide.
Statistical analysis. All experiments were done at least thrice. Data are expressed as mean ± SD and analyzed using a t test.
Results
Chronic TSH stimulation activates S6K1 in vivo. To identify the pathways that are effectively involved in the control of thyrocyte proliferation during prolonged TSH stimulation, we analyzed the activation status of candidate targets in the thyroids of age- and sex-matched wild-type mice that had been treated for 2 weeks with the goitrogenic compounds methimazole and sodium perchlorate ( 10).
The phosphorylation (and thus activation) of p44/42 mitogen-activated protein kinase (MAPK) was not altered by TSH stimulation, whereas, unexpectedly, the phosphorylation of CREB, an important downstream target of PKA, was severely reduced in goitrogen-treated mice, suggesting that proliferation upon chronic TSH exposure does not rely on signaling through these molecules ( Fig. 1A ). On the contrary, S6K1 phosphorylation was increased in TSH-exposed thyroids, and its functional activation was confirmed by the massive phosphorylation of one of its targets, ribosomal protein S6, as detected by immunohistochemistry ( Fig. 1A and B). Thus, S6K1 might play an active role in the proliferation of thyroid follicular cells upon prolonged TSH stimulation.
Chronic TSH stimulation activates S6K1, but not CREB or MAPK. A, Western blot analyses of total and phosphorylated CREB, S6K1, and ERK1/2 in protein extracts from control and goitrogen-treated mice. Blots were further normalized with β-actin (data not shown). B, immunohistochemical analyses of S6 phosphorylation in thyroid sections from control and goitrogen-treated mice.
TSH-induced hyperplasia is mediated by mTOR. Because S6K1 activation is primarily dependent on the mTOR kinase, we hypothesized that TSHR engagement might induce proliferation through the activation of the mTOR/S6K1 cascade.
To address this possibility, we treated cohorts of 6-week-old mice (n = 5–7) with goitrogens alone or in combination with the rapamycin derivative RAD001 (Everolimus), a specific inhibitor of mTOR, for 5 and 14 days. As expected, the thyroid weight of goitrogen-treated mice doubled over the 2-week treatment period. Histologic analysis revealed that the thyrocytes had lost their flat morphology and acquired hypertrophic, columnar features ( Fig. 2A and B ). RAD001 treatment alone, on the other hand, did not have any obvious effect on the thyroid size or histologic features ( Fig. 2A and B). Combined treatment with goitrogens and RAD001 resulted in only a minor reduction of the thyroid weight, and the morphologic features of the glands were indistinguishable from those of mice treated only with goitrogens, suggesting that the TSH-induced hypertrophy does not rely on mTOR activation ( Fig. 2A and B).
TSH proliferative signals converge on the mTOR pathway. A, graph showing the average thyroid weight, normalized by body weight, of control untreated, RAD001-treated, goitrogen-treated, and RAD001 + goitrogen-treated mice at the end of a 5-d or a 2-wk regimen. B, representative microphotographs showing the histology associated with each of the four treatment groups at 14 d. Pt, parathyroid. Original magnification, ×400. C, thyrocyte proliferative index in the four treatment groups, determined by Ki-67 immunohistochemistry at the end of a 5-d or a 2-wk regimen.
We then used Ki-67 immunostaining to measure the proliferation of the thyroid follicular cells. Control thyrocytes were characterized, as expected, by an extremely low proliferative index, which was not altered by RAD001 treatment ( Fig. 2C). Goitrogenic treatment significantly and progressively increased the number of proliferating thyrocytes (>13-fold increase at 14 days). Strikingly, the combined treatment with goitrogens and RAD001 drastically reduced the proliferative index, almost to control levels, at both time points. These results strongly suggest that a large part of the TSH-induced mitogenic signal is funneled through the mTOR pathway.
mTOR activation occurs independent of AKT. Activation of mTOR upon mitogenic stimuli is usually thought to be mediated by the AKT-dependent inhibition of TSC1/TSC2 complexes, and the subsequent activation of Rheb. In fact, TSH-dependent AKT activation has been shown in some (but not all) cell culture models of thyroid proliferation ( 9, 11, 12). At the same time, mTOR activation results in the phosphorylation of the S6 ribosomal protein, through the ability of mTOR to directly activate S6K1. Therefore, we used immunohistochemistry on thyroid sections from the four cohorts of mice to detect AKT phosphorylation on Ser473 and S6 phosphorylation on Ser235/236. S6 was heavily phosphorylated upon goitrogen exposure, thus further proving the direct involvement of the mTOR cascade downstream of TSH, and RAD001 treatment completely reverted S6 phosphorylation ( Fig. 3 ). However, and rather surprisingly, no change was detected in the activation of AKT ( Fig. 3).
mTOR pathway activation during chronic TSH exposure is independent of AKT. Representative microphotographs of the immunohistochemical detection of phosphorylated, active forms of AKT and S6 in the four treatment groups. Inset, a positive control for pAKT, represented by thyroid follicles with tissue-specific Pten inactivation.
To further analyze this proliferation-inducing signaling cascade activated by chronic TSH stimulation in vivo in the thyroid gland, we did Western blot analysis using antibodies against specific phosphorylation sites of relevant proteins in the AKT/mTOR/S6K pathway. mTOR activation upon in vivo TSH stimulation was once again shown by the phosphorylation of S6K1 on Thr389 and Thr421/Ser424 upon goitrogen treatment, which was totally abrogated by simultaneous RAD001 administration ( Fig. 4A ). Along the same line, ribosomal protein S6 was phosphorylated on Ser235/236 in goitrogen-treated thyroid extracts, but not in extracts from goitrogen plus RAD001-treated mice. However, phosphorylation of TSC2 on Thr1462 (an AKT target site), of AKT on Ser473 and Thr308, and of ERK1/2 on Thr202 and Tyr204, did not change upon any treatment, strongly suggesting that mTOR activation takes place independent of AKT or MAPK activation ( Fig. 4A).
A, Western blot analysis of the phosphorylation and activation status of relevant molecules in the mTOR pathway in protein extracts from control untreated (C), RAD001-treated (R), goitrogen-treated (G), and RAD001 + goitrogen-treated (GR) mice, at the end of the 2-wk regimen. B, Western blot analysis of the phosphorylation and activation status of S6K1 and AKT in protein extracts from primary thyrocytes pretreated with vehicle or RAD001 (R) and exposed to TSH (T) for 20 min.
Finally, we did Western blot analysis on primary cultures from 8-week-old mice grown in reduced serum for 48 h and stimulated with TSH for 20 min ( Fig. 4B). As in the in vivo system, TSH induced the phosphorylation of S6K1 on Thr421/Ser424, which was abolished by pretreatment with RAD001, but not the activation of AKT, as measured by the phosphorylation on Ser473. These data fully support our hypothesis that mTOR activation upon TSHR engagement is independent of AKT activation.
Discussion
The relative contribution of the different TSH-initiated cascades to the various aspects of thyroid physiology, differentiation, and growth is still largely unknown. Our data, obtained in vivo in a relevant model of thyroid proliferation under sustained TSH stimulation, support a novel picture in which the proliferative input upon TSH signaling is conveyed through the mTOR kinase. In fact, we show that continuous TSH stimulation, during a time window known to result in simultaneous continuous thyrocyte proliferation and gland hypertrophy ( 13), does not lead to MAPK activation or to activation of the CREB transcription factor, but rather results in CREB down-regulation. This latter finding is similar to the differential response of CREB phosphorylation in the mouse nucleus accumbens upon acute and chronic nicotine exposure ( 14), and might reflect a negative feedback mechanism consequent to chronic TSHR/PKA stimulation, as previously shown in human toxic adenomas ( 15).
Strikingly, our data show that it is the mTOR/S6K1/S6 axis that becomes hyperactive during chronic TSH stimulation, as detected using phosphorylation-specific antibodies. Furthermore, specific inhibition of mTOR activity through the rapamycin analogue RAD001 abolishes the phosphorylation of both S6K1 and its target ribosomal protein S6, simultaneously restoring virtually normal proliferation levels. These data unequivocally show the key role of mTOR in inducing thyrocyte proliferation downstream of TSHR. The involvement of S6K1 in TSH-dependent proliferation had been previously suggested by studies in dog and rat thyrocytes ( 6, 9), although the mechanisms of S6K1 activation downstream of TSH had not been defined.
Our data not only uncovers, for the first time, the essential role this cascade plays in thyroid proliferation in vivo, but also allows the uncoupling of two major TSH effects, the hypertrophic response of the thyroid follicles, and the burst of thyrocyte proliferation. Although the latter is almost completely under the control of mTOR signaling, likely through the increased translation of molecules such as cyclin D1 ( 16) and cyclin D3 ( 17), the hypertrophic response to TSH, in terms of both thyroid weight and cellular morphology, is virtually unchanged upon RAD001 treatment, strongly suggesting that pathways different from the mTOR cascade are responsible for the thyrocyte shape and size changes. This view is also supported by the notion that the hyperproliferative effects of chronic TSH exposure, in vivo, fade away after 2 to 3 weeks, whereas the hypertrophy persists for at least 14 weeks ( 13).
The major open question stemming from our findings, however, relates to the mechanisms involved in the TSH-dependent mTOR activation. The “classic” mechanism of mTOR activation upon growth factor receptor signaling relies on the AKT-mediated inhibition of TSC1/TSC2, which in turn, allows Rheb to activate mTOR. Our data, quite surprisingly, shows that AKT is not part of this pathway during chronic TSH exposure. We could not, in fact, show any changes in AKT phosphorylation at the two canonical activation sites under goitrogen exposure and under RAD001-mediated mTOR inhibition. These in vivo data support previous findings from in vitro models, in which TSH was not able to induce AKT phosphorylation in dog primary thyrocytes ( 9) and in rat FRTL-5 cells ( 12), but are in sharp contrast with the rat WRT cell model, in which AKT phosphorylation is induced by TSH exposure ( 11). Although species-specific differences have often been invoked to justify these discrepancies, we cannot exclude the possibility that some molecular characteristics of the in vitro systems may not faithfully reproduce the physiology of the thyroid response to TSH in vivo.
AKT-independent activation of mTOR is becoming increasingly recognized as a mechanism of control of cell proliferation. For example, it has been recently shown that prostaglandin F2α activates mTOR via MAPK, but not via AKT in luteal cells ( 18), whereas mTOR activation, independent of both AKT and MAPK, has been detected in B lymphocytes ( 19) and endometrial stromal cells ( 20). The dissection of the specific mechanisms involved in the TSH-dependent activation of mTOR is tightly linked to the identification of the key players downstream of TSHR, and those upstream of mTOR. The major limitation of our in vivo approach is represented by the impossibility to directly measure Rheb activation, thus precluding the identification of the exact point at which the TSHR signaling cascade intersects the mTOR-activating pathway. Primary ex vivo thyrocyte culture systems will be necessary to work out the details of the connection between TSHR engagement and mTOR activation.
In summary, our studies have defined the mTOR kinase as a critical effector of the proliferative signals initiated by TSH, through a mechanism that does not require activation of AKT and is independent of the hypertrophic effects of TSH in the thyroid follicular cells.
Acknowledgments
Grant support: Appropriation from the Commonwealth of Pennsylvania and a National Cancer Institute core grant to Fox Chase Cancer Center.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The authors thank Dr. A. Bellacosa for critical reading of the manuscript. We acknowledge the Transgenic, DNA Synthesis, Laboratory Animal, Cell Culture, and Histopathology Core Facilities of Fox Chase Cancer Center.
Footnotes
-
Note: C. Brewer and N. Yeager contributed equally to this work.
- Received July 3, 2007.
- Accepted July 16, 2007.
- ©2007 American Association for Cancer Research.
References
Volume 67, Issue 17
