Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
A novel role for thyroid-stimulating hormone: Up-regulation of hepatic 3-hydroxy-3-methyl-glutaryl-coenzyme a reductase expression through the cyclic adenosine monophosphate/protein kinase A/cyclic adenosine monophosphate–responsive element binding protein pathway†
Endocrinology, Provincial Hospital affiliated to Shandong University, Jinan, China
Institute of Endocrinology, Shandong Academy of Clinical Medicine; Jinan, China,
Department of Endocrinology, People's Hospital of Gansu Province
Potential conflict of interest: Nothing to report.
Elevated thyroid-stimulating hormone (TSH) and hypercholesterolemia commonly coexist, as typically seen in hypothyroidism, but there is no known mechanism directly linking the two. Here, we demonstrated that in liver cells, TSH promoted the expression of 3-hydroxy-3-methyl-glutaryl coenzyme A reductase (HMGCR), a rate-limiting enzyme in cholesterol synthesis, by acting on the TSH receptor in hepatocyte membranes and stimulating the cyclic adenosine monophosphate / protein kinase A / cyclic adenosine monophosphate–responsive element binding protein (cAMP/PKA/CREB) signaling system. In thyroidectomized rats, the production of endogenous thyroid hormone was eliminated and endogenous TSH was suppressed through pituitary suppression with constant administration of exogenous thyroid hormone, and hepatic HMGCR expression was increased by administration of exogenous TSH. These results suggested that TSH could up-regulate hepatic HMGCR expression, which indicated a potential mechanism for hypercholesterolemia involving direct action of TSH on the liver. (HEPATOLOGY 2010)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Hypothyroidism is well known to be associated with elevated serum TC, which can result in hypercholesterolemia.1, 2 The underlying mechanism is widely thought to be TH deficiency. However, elevation of serum TC has also been observed in patients with subclinical hypothyroidism (SCH), in which TSH is elevated but TH stays within its normal range.1, 3, 4 Thus, the development of hypercholesterolemia in SCH cannot be explained only by the role of TH. This raises the question of whether elevated TSH also plays a role in the development of hypercholesterolemia in hypothyroidism. Several clinical studies in recent years addressed this issue and showed a correlation in hypothyroidism between high serum cholesterol and high TSH levels, the latter being usually used as an indication of the severity of the hypothyroidism.5-7 In these studies, however, it was impossible to delineate a direct role of TSH as abnormal TH level (usually deficiency) was usually a coexisting factor. Thus, a molecular mechanism by which TSH might affect cholesterol level has never been established.
The TSHR is expressed in thyroid cells and plays a central role in up-regulating its function, including the synthesis of TH. Increasing data showed that the TSHR was also expressed in many nonthyroid tissues, and it might actually play a physiological role in these tissues.8, 9 We recently demonstrated that TSHR was expressed in hepatocytes and that stimulation of cultured liver cells with TSH increased the production of cAMP.10 These results suggested that the TSHR might play an important role in the regulation of liver function.
In the present study, we tested the novel hypothesis that TSH, by binding to TSHR on hepatocytes, plays an important role in cholesterol synthesis by the liver. Although HMGCR is found in virtually all tissues, it is most highly expressed in the liver and functions as a rate-limiting enzyme in cholesterol synthesis by the liver.11 Therefore, using both in vitro and in vivo experimental approaches, we specifically investigated whether TSH might regulate HMGCR expression by the liver.
cAMP, cyclic adenosine monophosphate; CREB, cyclic adenosine monophosphate responsive element binding protein; HMGCR, 3-hydroxy-3-methyl-glutaryl coenzyme A reductase; mRNA, messenger RNA; PKA, protein kinase A; RNAi, RNA interference; Sh, sham-operated; siRNA, small interfering RNA; TC, total cholesterol; TH, thyroid hormone; TSH, thyroid-stimulating hormone; Tx, thyroidectomized.
Materials and Methods
Details can be found in Supporting Materials and Methods.
Human normal liver cell line L-02 and murine liver cell line BNL.CL.2 (BNL) were obtained from the Type Culture Collection of the Chinese Academy of Sciences, Shanghai, China. Human primary hepatocytes were isolated from normal human liver tissues of subjects undergoing elective liver lobectomies or resection of smaller fragments for medical reasons (see Supporting Materials and Methods for detailed protocols). When treated with TSH or other reagents, cells were cultured in serum-free medium.
Male Wistar rats weighing between 180 and 200 g and 6-8 weeks old were obtained from Shandong University Animal Laboratory. Rats were divided into two groups: one group (n = 60) was surgically thyroidectomized (Tx), another group was sham-operated (Sh) control (n = 18). Tx rats were given subcutaneous injections of either T4 (n = 48, 8 μg/kg body weight daily, Sigma) or a corresponding volume of vehicle (n = 12). Then the T4-treated rats were divided into four subgroups (n = 12 for each group), which consistently received subcutaneous injection of T4 along with freshly prepared TSH at a dose of 0.05 IU/rat (0.15 IU/kg body weight), 0.3 IU/rat (1 IU/kg body weight), 1.5 IU/rat (5 IU/kg body weight) or corresponding volume of vehicle daily for 7 days.12, 13 Blood from animals was then obtained for analyses of serum T4, TSH, TC, calcium, phosphorus and liver function. In addition, livers from all animals were collected and immediately frozen for assay.
Quantitative Real-time PCR
qPCR was performed by using the primers listed in Supporting Table 1, according to a method as described previously.14
Western Blot Analysis and Immunofluorescence
See Supporting Materials and Methods.
Assay of HMGCR Activity
Hepatic microsomes were prepared as described by Honda.15 The method for the measurement of HMGCR activity was carried out as described previously.16
Determination of Intracellular cAMP and TC
See Supporting Materials and Methods.
Knockdown of TSHR by RNAi
RNAi candidate target sequences to human or mouse TSHR were designed (Supporting Table 1). L-02 cells were infected with lentivirus expression vectors (GeneChem, Shanghai) and BNL cells were transfected with transfection reagent (Dharmacon). Details can be found in Supporting Materials and Methods.
Plasmid Construction and Luciferase Reporter Assay
The -641 to +125 region containing CRE element of human HMGCR promoter was amplified from human genomic DNA template and inserted into PGL4.15 empty vector (Promega), named as pGL4-CRE. Mutated CRE binding site of this HMGCR promoter (TGACGTAG to TAAAAGGG) were inserted into the equivalent site of the pGL4.15 to generate the CRE mutant designated as pGL4-muCRE. After transfected with 0.2 μg pGL4-CRE or pGL4-muCRE for 16 hours, L-02 cells were devoid of serum and subsequently incubated with forskolin or TSH for another 12 hours. pRL-TK was used to normalize the luciferase activity. Cells were harvested and luciferase activities were measured using a dual-luciferase reporter assay system (Promega).
Electrophoretic Mobility Shift Assay (EMSA) and Chromatin Immunoprecipitation (ChIP)
Both assays were performed as previously described.17 Antibodies and primers listed in Supporting Materials and and Supporting Table 1.
Data were analyzed using SAS 9.1.3 and expressed as means ± standard deviations. Differences between means were compared using either unpaired Student t tests for two-group comparisons or one-way analysis of variance (ANOVA) (Dunnett's t or LSD test) for multiple comparisons. ANOVA (repeated measure) was performed to determine treatment effects of T4 and TSH on animal models. Differences were considered significant at P < 0.05.
TSHR Is Functional in Liver Cells
We previously demonstrated that TSHR expressed in liver cells, including human liver cells.10 Here, we took further steps to examine and demonstrate a functional coupling of the TSHR to the cAMP system in the cells. Treatment with TSH significantly stimulated cAMP production in liver cells over the control (Fig. 1; P < 0.001), which was similar to that induced by forskolin (an adenylyl cyclase [AC] activator). It is known that hepatocytes express cell-surface receptors for glucagons, which coupled to the AC/cAMP system.18 We found that the effect of TSH on cAMP was similar to that of glucagons in liver cells. However, CHO cells that did not express TSHR showed enhanced cAMP production in response to forskolin (P < 0.001) but not to TSH (Fig. 1B).
TSH Increased the HMGCR Levels in Liver Cells
HMGCR protein, messenger RNA (mRNA), and activity all observed a dose-dependent increase in L-02 cells following TSH stimulation for 48 hours (Fig. 2A). Moreover, the increase of HMGCR protein and mRNA level became evident at 24 hours after treatment with TSH, and with more pronounced effect at 48 hours (Fig. 2B). Similar results in HMGCR protein expression were also found in human primary hepatocytes and BNL cells after TSH treatment (Supporting Fig. 1).
TSH Increased the Production of TC in L-02 Cells
As LDL receptor (LDLR) is a key player in cholesterol metabolism, we compared the in vitro effects of T3 and TSH on the expression of LDLR. T3 stimulated LDLR protein expression in L-02 cells in a concentration-dependent manner (Fig. 2C). However, we did not see an obvious effect of TSH on LDLR expression, in striking contrast with the effect of TSH on the HMGCR expression. Furthermore, treatment of L-02 cells with increasing concentrations of TSH resulted in a concentration-dependent increase in intracellular TC. This pattern of TC changes was very similar to that of the change in HMGCR in response to TSH stimulation.
TSH-Stimulated Responses in Liver Cells were Dependent on the Presence of TSHR
We performed a series of experiments to investigate whether TSH induced HMGCR expression in liver cells via TSHR as TSH acts in thyroid gland.
To block TSHR, we used a monoclonal antibody (CS-17) with competitive antagonist properties against human TSHR.19, 20 The results showed that TSH-stimulated production of cAMP in L-02 cells and human primary hepatocytes cultured in the presence of CS-17 was significantly lower than that in cells cultured without CS-17 (P < 0.001) (Fig. 3A, upper). Moreover, both basal and TSH-stimulated HMGCR protein levels in L-02 cells were substantially reduced by CS-17 (Fig. 3A, lower).
We also used a lentivirus-based RNA interfere (RNAi) delivery system to knock down the expression of TSHR in L-02 cells. Fluorescent microscopic examination revealed that the efficiency of lentiviral infection was higher than 90% at 72 hours (Supporting Fig. 2). As shown in Fig. 3B, the expression of TSHR was significantly and specifically knocked down by RNAi. Correspondingly, TSH-stimulated cAMP levels, HMGCR protein and TC production were greatly diminished in cells infected with RNAi lentivirus. In contrast, in cells infected with negative control lentivirus (NS lentivirus), TSH could still increase cAMP levels, up-regulate HMGCR protein and enhance TC production. Treatment of cells with NS lentivirus or RNAi lentivirus alone had no effect on HMGCR protein expression.
In separate experiments, we used siRNA to knock down TSHR expression in BNL cells and achieved similar results to those in the L-02 cells with RNAi approach (Supporting Fig. 3).
The cAMP/PKA/CREB Pathway was Involved in TSH-Induced Up-Regulation of HMGCR
TSH-stimulated cAMP production in L-02 cells and human primary hepatocytes was significantly inhibited by treatment with AC inhibitor (SQ22536) (P < 0.001) (Fig. 4A). Similarly, the protein expression of HMGCR in L-02 cells stimulated by TSH was dramatically reduced by SQ22536 (Fig. 4A). These suggested that TSH increased HMGCR levels in liver cells through a cAMP-dependent pathway.
It was reported that the HMGCR promoter contained a cAMP-responsive element CRE.21, 22 We constructed a recombined luciferase reporter plasmid pGL4-CRE and transfected into L-02 cells. The significant increase in luciferase activity was detected upon TSH or forskolin treatment. After we mutated the CRE binding site of HMGCR promoter (pGL4-muCRE), we found neither forskolin nor TSH could up-regulate its luciferase activity, which strongly indicated that the CRE site was essential for TSH in regulation of HMGCR (Fig. 4B).
To assess whether TSH has any effect on DNA-binding activity of CREB with CRE locating HMGCR promoter, EMSA was performed. Results showed that CREB-DNA binding activity was specific because the band disappeared with an excess of unlabeled CRE, whereas the mutant failed to influence the bound. Specificity of the binding complex was further demonstrated using a CREB or phosphorylated CREB (Ser133 pCREB) antibody (Fig. 4C). This nuclear protein/DNA complex was more abundant when cells treated with TSH (Fig. 4C). To investigate whether PKA is also involved in increased CERB-DNA binding activity stimulated by TSH, PKA inhibitor H89 was added, and the faint gel bands were found. In addition, ChIP assay showed that TSH markedly increased pCREB binding capacity in comparison to the control (P < 0.001), whereas H89 dramatically down-regulated this activation by TSH (P = 0.019 versus control). Likewise, knockdown of TSHR by RNAi inhibited TSH-induced CREB activation (P = 0.002 versus TSH) (Fig. 4D).
Taken together, these results suggest that TSH-induced elevation of cellular cAMP levels activates PKA. PKA in turn phosphorylates and activates CREB, which transcriptionally activates HMGCR.
TSH Enhanced Hepatic HMGCR Protein Expression In Vivo
To further investigate the role of TSH in the regulation of HMGCR, we pursued in vivo studies in rats. Circulating T4 was reduced to an undetectable level whereas serum TSH was dramatically elevated, and this was accompanied by a significant increase in plasma TC (P = 0.041) in Tx rats compared with the Sh rats (Fig. 5A, Table 1). After treatment with T4, endogenous TSH levels in Tx rats were reduced to low levels. Moreover, administration of T4 to Tx rats reduced the elevated serum TC to levels similar to those observed in Sh rats. In addition, hepatic tissue proteins from Tx and Sh rats were analyzed for HMGCR and LDLR proteins, respectively. A significant increase (P = 0.004) in the HMGCR and a significant decrease (P = 0.038) in the LDLR in Tx rats relative to Sh animals were observed (Fig. 5B).
Table 1. Comparison of Various Serum Biochemical Parameters in Sh and Tx Rats (Mean ± Standard Deviation)
Tx (N = 60)
T4 and TSH (0.05 IU)
T4 and TSH (0.3 IU)
T4 and TSH (1.5 IU)
ALT, alanine aminotransferase; AST, aspartate aminotransferase; BW, body weight; TC, total cholesterol. T4 was given at constant doses (8 μg/kg BW daily); TSH was given at a dose of 0.05 IU/rat/day or 0.3 IU/rat/day or 1.5 IU/rat/day; Control was given a corresponding volume of vehicle.
We then administered exogenous TSH to these Tx rats at 0.05, 0.3, or 1.5 IU/rat daily for 7 days while they received daily T4, respectively. There was no significant difference in the serum T4 levels in the group of Tx rats receiving only exogenous T4 compared with the Tx rats receiving both exogenous T4 and TSH (P > 0.05), whereas the serum TSH levels statistically increased in the group of Tx rats receiving exogenous TSH compared with the Tx rats receiving only exogenous T4 (Fig. 6A, upper). Furthermore, we observed a dose-dependent increase in serum TC after administration of exogenous TSH, although this increase marginally failed to reach statistical significance (P > 0.05) when comparing the group of Tx rats receiving only exogenous T4 with the group of Tx rats receiving both exogenous T4 and TSH (Fig. 6A, upper). However, in the same group of Tx rats constantly receiving exogenous T4, a significant increase in serum TC was observed after TSH injection compared with before injection (Fig. 6A, lower). No significant difference in serum calcium, phosphorus, or liver function (alanine aminotransferase and aspartate aminotransferase) was observed among the different groups of animals.
Given the up-regulating effects of TSH on HMGCR in liver cells demonstrated in our in vitro experiments presented above, we tested the possibility that TSH plays a role in the regulation of hepatic HMGCR in vivo. Administration of T4 alone decreased hepatic HMGCR expression in Tx rats (Fig. 6B), likely due to normalization of the elevated endogenous TSH in Tx rats. Remarkably, in Tx rats consistently receiving T4, administration of exogenous TSH, particularly at the higher dose, significantly increased the protein level of hepatic HMGCR. In contrast, although the level of hepatic LDLR protein in Tx rats was increased by administration of T4, no further increase was observed after additional administration of exogenous TSH at either dose. These findings were consistent with the in vitro results in liver cells, as presented above, that TSH stimulated expression of HMGCR, but not LDLR. Furthermore, the changes of TC levels in liver tissue were similar to those of HMGCR in all groups of experimental rats (Fig. 6C). These suggested that TSH could increase hepatic TC levels by up-regulating HMGCR.
We have previously demonstrated the expression of TSHR protein in liver cells.10 In the present study, by showing its coupling to the intracellular cAMP system and the expression of HMGCR, we established the functionality of this receptor in liver cells. This was unequivocally proven by the abolishment of the effects of TSH in cells treated with specific TSHR monoclonal antibodies or lentiviral TSHR siRNA to silence the expression of TSHR.
In the present study, we demonstrated a significant increase in the expression of both mRNA and protein of HMGCR in response to TSH stimulation in hepatocytes. This effect of TSH was dose-dependent and time-dependent as well as TSHR-dependent. It should be noted that the TSH concentrations used in the present study were higher than that in normal people or patients with hypothyroidism, similar to the concentrations used for thyrocytes in culture23 or for nonthyrocytes in culture, such as 3T3-L1 preadipocytes24 and fibroblasts.25 The reason for using a lower concentration of TSH in human body is possibly the synergistic action of coexisting growth factors/cytokines such as IGF-1 to augment TSH signaling in vivo.23
The data presented strongly support the role of cAMP as a mediator of the stimulatory effects of TSH on HMGCR gene expression. However, there are some proteins that bind to the promoter for HMGCR which are thought to be responsible for transcriptional regulation. For example, insulin, a known activator of HMGCR, could enhance CREB transcriptional activity in HepG2 cells through the induction of CREB phosphorylation.26 Once CREB has been activated, it interacts efficiently with sterol regulatory element binding protein-2 to stimulate the transcription of the HMGCR gene in the presence of NF-Y.22 It is conceivable that TSH, through cAMP signal, could induce one or more such regulatory proteins to be actived in promoting reductase gene transcription. The data presented in this work demonstrated that TSH might induce HMGCR transcription at least through the cAMP/PKA/CREB pathway. This event requires that CREB becomes phosphorylated by PKA at Ser133 and acts at the major CRE within the HMGCR promoter region. Even though there was no further research for other regulatory proteins in the present study, our results also demostrated that activation of CREB by TSH in hepatocytes was found to contribute to increased gene expression of HMGCR.
A unique experimental approach in the present study was the use of surgically thyroidectomized rats that completely lost the ability to produce endogenous thyroid hormones and were subsequently treated with exogenous T4 to correct hypothyroidism and maintained a constant serum level of thyroid hormone as well as stably suppressed endogenous TSH through feedback from the pituitary gland. With this approach, we were able to alter the TSH levels in the body of the animal by administering exogenous TSH without altering the thyroid hormone levels which would otherwise have occurred through stimulation of the normal thyroid gland by exogenous TSH. Consequently, under these controlled conditions, we were able to test a sole role of TSH in cholesterol metabolism. As a result, we were not only able to demonstrate a role of TSH in up-regulating hepatic HMGCR expression in vivo but also a corresponding increase in serum TC.
In this study, thyroidectomy with resulting hypothyroidism itself caused elevated hepatic expression of HMGCR in rats. This is somewhat inconsistent with the results of Ness and Gertz, which showed lower expression of hepatic HMGCR in Tx rats.27 The explanation for this discrepancy might lie in differences in some of the experimental conditions, such as the duration of hypothyroidism and the types of foods (e.g., cholesterol contents) used to feed the animals. It is notable that in the studies of Ness and Gertz, the Tx animals were commercially obtained and likely had long-term hypothyroidism. Relatively long-term hypercholesterolemia that likely had occurred through other mechanisms, such as the TH deficiency-promoted down-regulation of LDLR in hepatocytes in such chronic hypothyroid conditions, could itself down-regulate HMGCR through a negative feedback mechanism. It is well known that a high level of serum cholesterol, such as that seen after intake of foods rich in cholesterol, can dramatically decrease HMGCR expression in liver.28
In contrast, the elevated hepatic HMGCR expression seen in our Tx animals occurred in a relative acute phase of hypothyroidism in which the positive effect of elevated TSH was probably quick and strong so that it overwhelmed the negative effect of the early and therefore still relatively mild hypercholesterolemia on the expression of hepatic HMGCR. This possibility is consistent with our finding that in these acutely hypothyroid Tx rats, administration of T4 quickly and efficiently reversed the rise in HMGCR presumably through suppression of TSH secretion of the pituitary gland. This was one reason that we administered a relatively high and constant dose of T4 to suppress the endogenous TSH to a low and stable level in Tx rats, so a quick and controlled change in TSH level in the body of the animal could be conveniently achieved by administering exogenous TSH to conclusively test a sole effect of TSH. The decrease in serum TC by administering T4 in Tx rats occurred through a dual mechanism involving a decrease in hepatic HMGCR expression through suppression of endogenous TSH as discussed above and an increase in hepatic LDLR expression as shown in the present study.
In summary, using a variety of unique in vitro and in vivo approaches, we demonstrated that TSH, by acting on the TSHR in liver cells, could up-regulate the expression of hepatic HMGCR through cAMP/PKA/CREB signal pathway. The results revealed a potential effect of TSH on cholesterol level by the liver and had possible pathological and clinical implications for the pathogenesis of hypercholesterolemia particularly that associated with hypothyroidism, which is a common human disease that is associated with elevated TSH.
The authors gratefully acknowledge Professor Basil Rapoport and Chunrong Chen for providing CS-17. We thank Zhu Chen, a member of the Chinese Academy of Science, for professional guidance on the subject. We also thank Professor Xiao Han for assistance in the EMSA experiment.