Nuclear Receptor Expression in Human Differentiated Thyroid Tumors

Abstract

Background:

Nuclear receptors (NRs) play a key role in endocrine signaling and metabolism and are important therapeutic targets in a number of hormone-dependent malignancies. Studies on the role of NRs in thyroid cancer are limited.

Objective:

The objective of the study was to examine systematically the expression of the 48 human NRs in a series of benign and malignant thyroid tissues. Within the papillary carcinoma cohort, we sought to determine if NR expression differed significantly by BRAF mutation status.

Patients and methods:

RNA was isolated from multinodular goiter (MNG; n=6), papillary carcinoma (PTC, n=14), follicular carcinoma (FC; n=5), and Hürthle cell carcinoma (HCC; n=7). The 48 human NRs were profiled in this panel by quantitative real time polymerase chain reaction. Protein expression for selected NRs (Rev-erbα and LXR-β) was examined by immunohistochemistry (IHC) on tissue microarrays comprising benign and malignant thyroid tissues.

Results:

Across all groups of benign and malignant thyroid tissue, there was prominent expression of LXR-β and ROR-γ. Key findings in PTC were marked overexpression of RXR-γ and Rev-erbα compared to MNG. Within the PTC cohort, when BRAF ^V600E tumors were compared with wild type BRAF, there was relative upregulation of RXR-γ and Rev-erbα and downregulation of AR, ERR-γ, and ROR-γ. In FC, EAR-2 was overexpressed, while PPAR-α and PPAR-δ were underexpressed compared to MNG. The NR expression profile of HCC was distinct, characterized by significant downregulation of a wide range of NRs. IHC for Rev-erbα and LXR-β localized protein expression to the tumor cells. Moderate to strong Rev-erbα immunostaining was seen in 22 out of 23 PTC, and, overall, staining was stronger than in the benign group.

Conclusions:

These results represent the first systematic examination of NR expression in thyroid cancer. Our finding of tumor-specific patterns of NR expression, as well as significant differences in NR expression between BRAF ^V600E and wild type BRAF PTC, provides a basis for further mechanistic studies and highlights potential novel therapeutic targets for this malignancy.

Introduction

Thyroid cancer is the most common endocrine malignancy. There is a strong female preponderance, and epidemiological evidence suggests that the overall incidence of thyroid cancer is increasing (1).

Differentiated epithelial thyroid cancer includes papillary carcinoma (PTC) and follicular carcinoma (FC). Hürthle cell carcinoma (HCC) or oncocytic follicular carcinoma has been considered a subtype of FC, although these tumors have distinct morphological features and are characterized by underlying mitochondrial alterations (2). The majority of differentiated thyroid cancers are effectively treated by thyroidectomy with or without radioactive iodine ablation. In a minority of cases, there is persistent disease, which may result in considerable morbidity and, in some cases, mortality. Advanced thyroid cancer is often refractory to radioiodine therapy, and treatment options are limited, with most tumors responding poorly to conventional chemotherapy (3). Tyrosine kinase inhibitors have shown promise. However, only a subset of patients respond, and the duration of response is variable (4). Consequently, there is a need for new therapies targeted at the underlying biology of thyroid tumors.

The nuclear receptor (NR) superfamily members include ligand-dependent transcription factors, as well as apparently ligand-independent orphan nuclear receptors. NRs play a key role in endocrine signaling in metabolism and hormone-dependent malignancies. There have only been limited studies on the role of NRs and their expression in normal thyroid tissue and thyroid malignancy. Estrogen receptor-α and -β expression in normal and malignant thyroid tissue has been examined, although results are conflicting (5 –8). Studies using thyroid cancer cell lines have suggested that estrogen is pro-proliferative and may promote tumor progression (9,10). In vitro and in vivo studies of retinoids have been conducted in thyroid cancer on the basis of purported effects on proliferation and differentiation (11 –16); the expression of retinoid receptors has been studied in this context (17,18). A potential tumor suppressor role of thyroid hormone receptor in thyroid malignancy has been reported in a mouse model of FC (19). Little is known about the role of other NRs in thyroid cancer, nor has there been a systematic expression profiling of the entire NR superfamily. Although such data might in principle be obtained from microarray data sets, the relatively low abundance of transcription factors means that NRs are often not captured in microarray analysis.

Therefore, in order to evaluate the expression of NRs in differentiated thyroid cancer systematically, we used the commercial ABI TaqMan^® Low-Density Nuclear Receptor Gene Signature Array (TLDA; Applied Biosystems. Foster City, CA; Cat. no.: 4379961) to screen cohorts of PTC, FC, HCC, and multinodular goiter (MNG). MNG was used as a benign comparator on the basis that nodule formation in these glands is distinct (20 –23).

We further subdivided the PTC cohort according to the presence of the BRAF ^V600E mutation. BRAF is a member of the RET/RAS/RAF/MAPK signaling pathway. The BRAF ^V600E mutation is exclusive to PTC and is the most common somatic mutation (40–60%) in this tumor. Evidence supports its role in the pathogenesis of PTC (24).

This data set corroborates and extends previous studies on individual NRs in thyroid cancer. The distinct tumor-specific NR gene expression profiles in these tumors provide a basis for further mechanistic studies and suggests potential novel therapeutic targets for the treatment of thyroid malignancy.

Materials and Methods

Isolation of RNA from tissue

Prior to RNA extraction, a representative section of each malignant thyroid tumor was reviewed by a pathologist and scored for the percentage of tissue components such as tumor, stroma, and inflammatory infiltrate. Only those tissues in which the representative sample contained at least 70% tumor component were included in the study. MNGs were sampled randomly.

RNA was extracted from frozen tissue using the PureLink RNA Mini Kit (Invitrogen, Carlsbad, CA). The RNA quality was checked by electrophoresis using a Bio-Rad Experion™ automated electrophoresis system (Bio Rad Laboratories, Hercules, CA). RNA was isolated from PTC (n=14), HCC (n=7), FC (n=5), and MNG (n=6). The details of the individual tumors and the BRAF mutation status of PTC are shown in Table 1. The collection and use of this tissue were approved by the Research and Ethics Committee of Box Hill Hospital, Box Hill, Australia. Tissue was accessed through the Victorian Cancer Biobank, and all patients gave written informed consent for the collection of tissue.

Table 1.

Clinical Information

PTC
Sample	Age	Sex	BRAF^V600E mutation	Lymph mode metastases	Invasion	Additional comments
PTC1	30	F	Y	Y	No
PTC2	64	M	Y	N	No
PTC3	71	F	Y	NS	Focal capsular invasion	Diffuse tumor R lobe
PTC4	63	F	N	Y	Vascular invasion	Poorly differentiated
PTC5	35	F	N	Y	Invades adjacent muscle
PTC6	50	F	Y	Y	Extrathyroidal extension	Recurrent PTC
PTC7	81	F	N	N	No
PTC8	72	F	Y	N	Extrathyroidal extension	Recurrent PTC
PTC9	37	F	N	NS	No
PTC10	55	F	Y	Y	No
PTC11	53	F	N	NS	No	Follicular variant
PTC12	38	F	Y	Y	Vascular and capsular invasion
PTC13	22	F	N	N	No	Follicular variant
PTC14	14	F	N	Y	Focal lymphovascular invasion and extrathyroidal extension	Follicular variant

HCC and FC				MNG
Sample	Age	Sex	Invasion	Sample	Age	Sex
HCC1	80	M	Focal vascular and capsular invasion	MNG1	59	F
HCC2	79	M	Focal vascular and capsular invasion	MNG2	75	F
HCC3	85	F	Focal vascular and capsular invasion	MNG3	46	F
HCC4	73	F	Focal vascular and capsular invasion	MNG4	68	F
HCC5	46	F	Focal capsular invasion	MNG5	64	F
HCC6	82	F	Widely invasive: vascular + perithyroid adipose	MNG6	59	F
HCC7	55	F	Focal capsular invasion
FC1	54	F	Poorly differentiated, vertebral metastases
FC2	76	F	Extensive vascular/extracapsular invasion
FC3	76	F	Extensive vascular invasion
FC4	57	F	Limited vascular invasion
FC5	80	M	Widely invasive, vertebral metastases

PTC, papillary thyroid carcinoma; HCC, Hürthle cell carcinoma; FC, follicular carcinoma; MNG, multinodular goiter; F, female; M, male; Y, yes; N, no; NS, not sampled.

BRAF mutation detection

Screening for BRAF mutations was by real time polymerase chain reaction (RT-PCR) as described previously (25) (see Supplementary Data, available online at www.liebertpub.com/thy).

TLDA

As in previous studies of NR gene expression (26,27), commercial microfluidic cards, the TLDAs that contain an exclusive set of TaqMan Gene Expression Assays for the 48 NRs and 16 internal controls were used to profile gene expression (Supplementary Data). These controls span the relative abundance/Ct range of the genes on the card.

For each sample, 1.5 μg of total RNA was reverse transcribed using random hexamers with SuperScript III reverse transcriptase (Invitrogen) in a total volume of 45 μL. A total of 100 μL reaction mixture containing 50 μL cDNA template (333 ng) in RNase-free water and an equal volume of TaqMan universal master mix (Applied Biosystems) was added to each TLDA fill reservoir. Four reservoirs per sample were filled. The TLDA includes all NRs and endogenous controls in triplicate. After sealing the plate, it was run on an ABI 7900HT Real Time instrument (Applied Biosystems).

Statistical analysis

The TLDAs were analyzed using the relative quantification method of ΔCt (Fig. 1). The geNorm software imbedded within the ABI/Intergromics StatMiner v4.1 software package was used to compute least expression variation and select the most appropriate, stable, and robust combination of internal control genes with which to normalize the expression data (against the mean of the most stable controls). The expression data were normalized to the median of HPRT1, PPIA, and TFRC.

FIG. 1.

Nuclear receptor gene expression in multinodular goiter (MNG; n=6); papillary carcinoma (PTC; n=14); follicular carcinoma (FC; n=5); Hürthle cell carcinoma (HCC; n=7). Samples with a CT value of ≥35 were given a value of 0. Data are presented as the relative expression normalized to the median of the geNorm selected controls of HPRT1, PPIA, and TFRC. The bar graphs represent mean±standard error of the mean (SEM) for each of the tumor types. Each tumor sample was run in triplicate.

Significant changes in expression relative to reference samples or an FDA approved universal human tumor RNA sample (Stratagene, La Jolla, CA), were analyzed using the ABI/Integromics StatMiner software package as previously done (26,27); see Supplementary Data.

Tissue microarray

Formalin-fixed, paraffin-embedded blocks routinely prepared from surgical specimens were selected. Representative areas containing tumor were identified by a pathologist. Duplicate tissue cores with a diameter of 1 mm were taken from each specimen (Beecher Instruments, Silver Spring, MD). The following tissues were studied: benign (normal thyroid, n=2; MNG, n=6); PTC (BRAF wild type, n=9; BRAF ^V600E, n=14); FC (n=5); and HCC (n=5).

Immunohistochemistry

Three micrometer tissue sections were cut from each arrayed paraffin block and prepared on pathological slides. Prior to staining, the sections were dewaxed and rehydrated in graded ethanol washes. After microwave antigen retrieval, endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 10 minutes. The sections were incubated with anti Rev-erbα polyclonal antibody, dilution 1:100 (Sigma Aldrich, St. Louis, MO; Cat. no.: HPA007935) and anti LXR-α/β polyclonal antibody, dilution 1:100 (Lifespan Biosciences, Seattle, WA; Cat. no.: LS-B262). Sections were incubated at room temperature for 60 minutes with the primary antibody dissolved in EnVision FLEX Antibody Diluent (Dako, Glostrup, Denmark). Subsequently, the sections were incubated for 30 minutes with Envision Link HRP (Rabbit/Mouse) reagent (Dako). Diaminobenzidine (DAB) was used as a substrate for staining development, and slides were counterstained with hematoxylin. Negative controls were stained with rabbit polyclonal IgG (Abcam, Cambridge, United Kingdom; Cat. no.: 27472) at the same concentration as the primary antibodies.

Immunohistochemical interpretation

Immunohistochemical stains for Rev-erbα were scored, blinded to the histopathologic diagnosis and BRAF mutation status of the tumors. Staining intensity in the epithelial cells was scored as 0 (no staining); 1+ (faint), 2+ (moderate), or 3+ (strong).

Results

Details of the thyroid tissues examined and the BRAF status of PTC are shown in Table 1. The relative expression of each NR normalized against software-selected geNorm controls for MNG, PTC, FC, and HCC is shown in Figure 1. We use the same names as the Nuclear Receptor Signaling Atlas (NURSA), the HUGO Gene Nomenclature Committee (HGNC), and the International Union of Basic and Clinical Pharmacology (IUPHAR) for the NRs; the systematic numbering nomenclature can be found in Supplementary Table S1.

In MNG, the highest levels of expression were seen for Nur77 and the glucocorticoid receptor (GR), with abundant expression also seen for liver X receptor-β (LXR-β), retinoid X receptor-α (RXR-α), retinoic acid receptor (RAR)–related orphan receptor-γ (ROR-γ), and peroxisome proliferator-activated receptor-δ (PPAR-δ). The NR expression profile of MNG was compared with the NURSA mouse thyroid data set (28). Concordance in relative expression was seen for the majority of NRs, with several exceptions. Of the six most abundantly expressed NRs in each group, concordance was seen for LXR-β, RXR-α, ROR-γ, GR, thyroid hormone receptor-α (TR-α), and PPAR-δ. The expression profiles of three NRs were dichotomous, including Nur77 (abundant in MNG only) and vitamin D receptor (VDR) and RXR-β (abundant in mouse thyroid only).

Considerable variability in NR expression was observed in both PTC and FC, with very low or no expression seen for a number of NRs, while other receptors showed relatively abundant expression. The PTC group was remarkable for high expression levels of Rev-erbα and RXR-γ. As with MNG, abundant expression was also seen for LXR-β, RXR-α, ROR-γ, and PPAR-δ.

In FC, the highest expression levels were seen for LXR-β, with abundant expression also seen for ROR-γ, Nur77, RXR-α, RXR-γ, and v-erbA-related receptor (EAR-2).

The HCC cluster was notable for low relative expression of the majority of NRs. Within this group, LXR-β and Rev-erbα showed higher relative expression compared to the other NRs.

Figure 2 shows comparisons of relative expression between the three tumor groups (PTC, FC, HCC) and MNG (Fig. 2A–C) normalized to software-selected geNorm controls (unique to each comparison). These data are expressed as a fold change (either increased or decreased expression relative to baseline) on a logarithmic scale. Figure 2D illustrates the comparison between BRAF mutation–positive and wild type BRAF PTC. All of the comparisons shown in Figure 2 represent statistically significant differences in NR expression (p<0.05), recognized as valid by the StatMiner software. Further data showing nonsignificant differences in NR expression can be found in the Supplementary Data.

FIG. 2.

Genes identified as significantly different in StatMiner analysis of (A) PTC versus MNG, (B) FC versus MNG, (C) HCC versus MNG, and (D) papillary BRAF mutation–positive carcinoma (n=7) versus papillary wild type (WT) BRAF carcinoma (n=7). Data are presented as relative quantification (Log₁₀) utilizing MNG (A–C) or papillary BRAF WT carcinoma (D) as the reference/calibrator normalized to geNorm controls. A unique set of geNorm controls was utilized for each comparison. For PTC versus MNG: HPRT1, PPIA, and YWHAZ; for FC versus MNG: PPIA, TFRC, and YWHAZ; for HCC versus MNG: HPRT1, PGK1, and PPIA; for BRAF ^V600E versus BRAF WT: GUSB, IPO8, and YWHAZ. Only those NRs with a significant difference in expression (p<0.05) and recognized as valid by the StatMiner software are shown. The solid bars represent results in which the Ct values for the gene of interest were <35 in both the target and calibrator samples. The gray and open bars represent results in which the calibrator was not detected and the target was not detected respectively.

Key findings in PTC were marked overexpression of RXR-γ and Rev-erbα compared to MNG. Within the PTC cohort, when BRAF ^V600E tumors were compared with wild type BRAF, there was relative upregulation of RXR-γ and Rev-erbα, and downregulation of AR, ERR-γ, and ROR-γ. In FC, EAR-2 was overexpressed, while PPAR-α and PPAR-δ were underexpressed compared to MNG. The NR expression profile of HCC was distinct, characterized by the significantly decreased expression of a wide range of NRs. Although the short heterodimer partner (SHP) gene appears relatively upregulated in HCC, absolute levels are low in HCC and negligible in the other groups, so this finding is unlikely to be biologically significant.

The individual tumor levels for the five genes whose expression was significantly different (p<0.05) between BRAF mutation–positive and wild type BRAF PTC (Fig. 2D) are shown using scatter plots (Fig. 3). Of these genes, RXR-γ expression segregates most distinctly according to BRAF mutation status, with only one outlier in each group. Expression of Rev-erbα was relatively low in all the wild type BRAF PTC, whereas a subset of BRAF mutation–positive PTC had substantially higher levels, and one tumor had markedly elevated expression.

FIG. 3.

Scatter plot analysis showing individual tumor levels for the five significantly different (p<0.05) genes in the papillary BRAF mutation–positive carcinoma versus papillary BRAF wild type carcinoma comparison as identified by StatMiner (Fig. 2D). The middle line represents the mean and the lines below and above delineate the standard error of the mean.

Immunohistochemistry (IHC) for Rev-erbα and LXR-β localized protein expression to the epithelial cells and, unexpectedly, expression for both NRs was predominantly cytoplasmic in the malignant tumors. Figure 4 demonstrates immunostaining for Rev-erbα (Fig. 4A,B) and LXR-β (Fig. 4C,D) in a representative BRAF mutation–positive PTC. While LXR-β mRNA expression was abundant across all types of thyroid tissue, significant differences in expression were observed for Rev-erbα between PTC and MNG and between BRAF ^V600E PTC and wild type BRAF PTC. Therefore, the intensity of staining for Rev-erbα protein in a range of thyroid tissues was scored, and the results are shown in Table 2. Overall, the intensity of staining was stronger in malignant tumors than in the benign group. Moderate to strong intensity (2+/3+) staining was seen in all but one PTC. Strong intensity staining was seen in 8 out of 14 BRAF ^V600E PTC (57%) and four out of nine wild type BRAF PTC (44%). Staining at 3+ was also seen in three out of five HCC. The relationship between mRNA levels and protein expression was examined, and a positive correlation was seen (data not shown).

FIG. 4.

Representative section of BRAF mutation–positive papillary carcinoma. (A) Low power view showing diffuse homogenous 3+ immunostaining for Rev-erbα with negative control (inset). (B) High power view showing 3+ immunostaining for Rev-erbα. (C) Low power view showing diffuse homogenous 2+ immunostaining for LXR-β with negative control (inset). (D) High power view showing 2+ immunostaining for LXR-β.

Table 2.

Intensity of Immunohistochemistry Staining for Rev-erbα by Tumor Subtype

IHC staining	1+	2+	3+
Benign (n=8)	4	4	0
PTC BRAF wild type (n=9)	1	4	4
PTC BRAF ^V600E (n=14)	0	6	8
FC (n=5)	0	4	1
HCC (n=5)	1	1	3

Staining intensity in the epithelial cells was scored as 0 (no staining), 1+ (faint), 2+ (moderate), or 3+ (strong).

IHC, immunohistochemistry.

An unsupervised hierarchical cluster analysis was performed, and the results are presented as a heat map (Supplementary Fig. S1). Within the PTC group, clustering was observed for subsets of tumors, including five out of seven BRAF mutation–positive PTC (seen at far left of heat map). HCC also demonstrate some clustering while MNG and FC are distributed amongst the other groups without evidence of clustering.

Individual boxplot analyses of the 48 NRs showing comparative expression levels in PTC, FC, HCC, and MNG are presented in Supplementary Figure S2. NRs of particular interest in this comparative analysis include RXR-γ, which shows the highest expression levels in PTC, lesser but still relatively abundant levels in FC, low levels of expression in MNG, and virtually no detectable levels in HCC. RAR-β shows the highest expression levels in MNG, with reduced levels in the malignant groups (decreasing expression is seen in the order FC, PTC, HCC). PPAR-δ was comparatively abundant in MNG and PTC, whereas lower levels were seen in FC with expression further reduced in HCC.

NR expression levels in PTC, FC, HCC, and MNG were compared to a tumor reference panel (Supplementary Fig. S3). Consistent across the thyroid tumor subtypes is significantly increased expression of the mineralocorticoid receptor (MR) and ROR-γ and reduced expression of ROR-β relative to the tumor reference panel. Increased relative expression of RXR-γ is seen in PTC, and analysis by BRAF status again reveals that this increase is predominantly in the BRAF mutation–positive group. Rev-erbα expression is significantly increased only in the BRAF ^V600E group.

Discussion

In this study, we have systematically determined the expression levels of the 48 NRs in cohorts of PTC, FC, HCC, and MNG. Although there are a limited number of samples in each group, the standard errors for the majority of NR expression results are small, reflecting the relative homogeneity of the cohorts.

Given the potential limitations of using MNG as the benign comparator, the NR profile of this subgroup was compared to the NURSA mouse thyroid data set. Concordance was seen for the majority of abundantly expressed NRs. LXR-β was among the most highly expressed NRs in mouse thyroid and MNG, as well as the three tumor subtypes. IHC localized protein expression to the epithelial cells. In light of these findings, a possible role for LXR-β in thyroid hormone biosynthesis warrants exploration. ROR-γ was abundant in mouse thyroid and MNG, and relatively abundant in PTC and FC, but with lower levels in HCC. The significance of ROR-γ expression in thyroid tissue is uncertain, given its close link to inflammation (29). Discordant expression was seen for three NRs, which likely reflects species differences. However, it is plausible that some disparities are due to differences between normal thyroid and multinodular tissue. Nur77 was highly expressed in MNG, but levels were low in mouse thyroid, which may reflect its known role in regulating cellular proliferation. RXR-α and VDR were abundantly expressed in mouse thyroid but not in MNG. These NRs are also linked to cellular proliferation and/or differentiation, which might possibly account for the differences.

Significant differences in NR expression were seen between all types of thyroid malignancy and benign thyroid tissue. Of particular interest, significant differences were noted between BRAF mutation–positive and wild type BRAF PTC in terms of both upregulation and downregulation of several NRs.

BRAF mutation positivity has been linked to a number of molecular events in PTC, including upregulation of tumor-promoting factors such as VEGF and NF-ĸB, silencing of tumor suppressor genes, and suppression of iodine metabolism (24). Gene expression studies in PTC and FC have noted variability in gene expression according to underlying mutation status (30 –32). In particular, Rusinek et al. (31) reported profound differences in gene expression signatures between BRAF mutation–positive and BRAF mutation–negative PTC. A direct comparison of our results with data from gene array studies is hampered by a number of factors, including the sensitivity of microarrays to detect low abundance genes and issues related to oligonucleotide probe specificity (33,34). However, when added to the previous observations, our findings further support the notion of significant pathobiological differences between PTC that carry and those that lack the BRAF ^V600E mutation.

PTC

Key findings in PTC were significant upregulation of RXR-γ and Rev-erbα, and downregulation of ERR-γ compared to benign controls. When analyzed by BRAF mutation status, significant differences in expression of these NRs were noted between BRAF mutation–positive and wild type BRAF PTC.

RARs and RXRs

Retinoids are vitamin A derivatives. They play important roles in cell growth and differentiation through RARs and RXRs. Each has three major subtypes (-α, -β, -γ). RXRs appear to function primarily as heterodimer partners with other NRs, including PPAR, TR, and RAR.

Several studies have examined expression levels of RXR and RAR isoforms in thyroid tumors by RT-PCR or IHC. Most but not all studies have reported increased levels of RXR-γ and reduced levels of RAR-β in PTC (17,35 –38). In a microarray study examining genes differentially expressed between PTC and normal tissue, RXR-γ was among the most highly upregulated genes in PTC (39), consistent with our finding of markedly increased expression of RXR-γ in PTC (more than 10-fold) compared to benign controls. We have further shown that expression levels in BRAF mutation–positive PTC are significantly higher than in wild type BRAF PTC. A nonsignificant trend toward reduced RAR-β expression was observed in the PTC group. This reduction in expression was significant, however, when only BRAF mutation–positive PTC were compared to MNG.

RAR-β appears to play an important role in mediating the growth inhibitory effects of retinoids in various cancer cells (40,41). It has been suggested that aberrant expression or loss of RAR-β can contribute to retinoid resistance, implying that it might act as a tumor suppressor (42). A possible mechanistic association between BRAF mutation positivity and reduced RAR-β2 expression has been suggested via methylation-induced gene silencing (43).

The selective upregulation of RXR-γ in PTC is intriguing, although its significance is difficult to determine, since RXRs are promiscuous heterodimer partners, as well as having the ability to form homodimers. A study of RXR-γ knockout mice suggested a role for RXR-γ in regulating thyroid hormone levels and metabolic rate possibly through the hypothalamic–pituitary–thyroid axis (44).

The differential expression profile of retinoid receptors in BRAF mutation–positive compared to wild type BRAF PTC is a significant finding. It may have therapeutic implications, as preclinical studies have highlighted the importance of different receptor isoforms in the cellular response to retinoids (12,17). Whether the observed alteration in expression is mediated by downstream effectors of the MAPK pathway or via cross-talk with other signaling pathways remains to be elucidated.

Rev-erbα

Rev-erbα and β are widely expressed orphan NR. They are part of the network of “clock genes” and are involved in the integration of circadian rhythms and metabolic functions, including lipid and glucose homeostasis (45,46). Rev-erbα is a target gene of PPAR-γ and is involved in adipocyte differentiation via PPAR-γ (47). Rev-erbα has also been shown to play a role in inflammation in vascular smooth muscle cells with overexpression resulting in upregulation of NF-ĸB responsive genes such as IL-6 and COX-2 (48). A recent study identified Rev-erbα and PPAR-γ binding protein (PBP) as important transcriptional regulators in a subset of aggressive (ERBB2-positive) breast tumors (49). Rev-erbα and PBP were shown to upregulate several genes in the de novo fatty acid synthesis network, contributing to abnormal cellular energy metabolism. Inhibition of Rev-erbα and PBP resulted in apoptosis of the breast cancer cells.

In our study, Rev-erbα expression was significantly higher in PTC compared with MNG. Expression levels in FC were low and comparable to MNG, whereas HCC showed slightly higher expression. Particularly noteworthy was significant overexpression of Rev-erbα in BRAF mutation–positive versus wild type BRAF PTC. IHC for Rev-erbα protein demonstrated moderate to strong intensity staining in all but one PTC. A higher percentage of tumors in the BRAF mutation–positive group showed strong staining, and a correlation between mRNA levels and protein expression was observed. However, the difference between the groups was less marked than for gene expression, and interpretation is limited by the semi-quantitative nature of IHC scoring.

Protein expression for both Rev-erbα and LXR-β localized to the epithelial cells across all the thyroid tissues studied with little staining in the intervening stroma. Both nuclear and cytoplasmic staining was observed for both NRs, although cytoplasmic staining was predominant in some of the benign tissues and in the majority of malignant tumors. Subcellular localization of NRs is challenging on a number of accounts. Nucleocytoplasmic shuttling of NRs may be altered by a variety of stimuli in the cellular milieu (50). In addition, it is increasingly recognized that NRs can have multiple isoforms that are structurally very similar. While the NR isoforms may share identical antibody epitopes, these protein variants may have alternate subcellular localization and/or functional properties (51). A tendency toward increased cytoplasmic and reduced nuclear staining in PTC compared with MNG has been previously reported for ER-β (7). Predominant cytoplasmic staining for a number of NRs has also been reported in other malignancies (52 –54), and in some cases, a correlation with prognosis was found, suggesting that tumor-related factors might alter the subcellular localization of the proteins.

Increased expression of Rev-erbα in BRAF ^V600E PTC may be a consequence of constitutive MAPK pathway activation, as regulation of Rev-erbα by this pathway has been previously demonstrated in pancreatic beta cells (55). Cross-talk between Rev-erbα and the NF-ĸB pathway is also potentially important, as NF-ĸB signaling has been increasingly implicated in the development and progression of malignancy, including thyroid cancer (56).

The fact that markedly increased Rev-erbα expression was seen in a subset of BRAF mutation–positive PTC but not in wild type BRAF PTC suggests that BRAF mutation with consequent MAPK activation may be necessary but not in itself sufficient to drive the overexpression of this NR. Further studies are required to identify other factors that might impact on Rev-erbα expression in thyroid tumors as well as downstream consequences of its overexpression.

FC

The NR profile of FC bears similarities to that of PTC, particularly in relation to the pattern of retinoid receptor expression and the relative downregulation of a number of receptors (such as ERR-γ and GR) compared to MNG. These similarities may reflect the activation of common signaling pathways such as MAPK and possibly shared metabolic changes at the cellular level. Unique to FC, however, was relative downregulation of PPAR-α and PPAR-δ and upregulation of EAR-2 compared to MNG.

PPAR

PPARs are divided into three main subtypes (PPAR-α, -β, and -γ), which are encoded by separate genes with different patterns of expression. They bind ligands such as fatty acids and also act in concert with other NRs and coactivators to regulate gene transcription. PPARs have been implicated in multiple physiological processes, including cellular differentiation and metabolism as well as in tumorogenesis (57).

Zeng et al. (58) reported high expression levels of PPAR-δ by immunohistochemistry in both benign and malignant thyroid tumors, and found a correlation between PPAR-δ expression and the marker of proliferation, Ki-67. Functional studies in primary thyroid cell cultures (58) and in some cancer models (59,60) have suggested a pro-proliferative role for PPAR-δ, whereas other studies have demonstrated antiproliferative effects (61,62). This raises the possibility that the response to PPAR-δ may vary by cell type or depend on other ligands and growth factors in the cellular milieu.

In our NR profiling, PPAR-δ was the predominant PPAR isoform in all tissue types. It was abundantly expressed in MNG and PTC with lower levels in FC and HCC. In contrast to Zeng et al., we did not find significantly increased PPAR-δ expression in malignant tissue versus benign. In fact, along with PPAR-α, PPAR-δ expression levels were significantly lower in FC compared with MNG. This discrepancy may relate to differences in methodology used and/or the number and type of samples studied.

Modulation of PPAR has been investigated in thyroid cancer, primarily in trials of PPAR-γ agonists as therapy for radioiodine resistant tumors. The downregulation of specific PPAR isoforms in FC may have implications for therapy, and a better understanding of the isoform specific roles of the PPARs is needed.

EAR-2

EAR-2 is an orphan nuclear receptor. It is a member of the chicken ovalbumin upstream promoter-transcription factor subfamily. The functions of EAR-2 are not well understood, although it has been shown to coregulate thyroid hormone receptor function negatively (63). Recent publications have identified overexpression of EAR-2 in breast (27) and colon cancer (64). Functional studies in models of colon cancer (64) and leukemia (65) have suggested a role for EAR-2 in enhancing cell survivability and promoting cellular proliferation.

EAR-2 expression was significantly higher in FC compared to MNG. The finding of elevated EAR-2 levels in FC but not PTC underscores the distinct molecular pathogenesis of these two epithelial cancers. It all suggests that, at least in FC, EAR-2 may be a therapeutic target.

HCC

The key finding in HCC was significant downregulation of a wide range of NRs compared to MNG (with levels up to 10-fold lower for some NR). HCC are often considered a subtype of FC and, indeed, a number of the NRs downregulated in HCC were similarly downregulated in FC, including: ERR-γ, GR, PPAR-α, PPAR-δ, and TR-α. However, an additional 14 NRs were uniquely downregulated in HCC, including members of the retinoid family, testicular receptors, NGF1B, COUP-TF2, and LRH-1.

HCC are characterized by mitochondrial alterations, including mitochondrial gene deletions and mutations, which lead to energy production defects in Hürthle cell tumors (66). Mitochondrial proliferation associated with the expression of factors involved in mitochondrial biogenesis may be a compensatory mechanism for such defects. In a recent study examining mutation status, chromosomal copy number, and gene expression (67), HCC were found to have a unique profile that differed significantly from PTC and FTC. A retrospective clinical study involving a large cohort of patients found that HCC was associated with lower overall and disease-specific survival than other forms of differentiated thyroid cancer (68). Taken together, this all supports the notion that HCC are governed by unique tumor biology, and further investigation into the mechanisms for this, including the role of NRs, are needed.

Conclusion

Our study of NR expression in thyroid tumors corroborates and extends previous findings on individual NRs in thyroid cancer. We have shown that the pattern of NR expression varies both by tumor type and, in PTC, by BRAF mutation status. Two of the NRs differentially expressed in BRAF mutation–positive tumors—Rev-erbα and ERR-γ—are increasingly recognized as modulators of metabolism. Recent evidence suggests that a shift to glycolytic metabolism may occur in BRAF mutation–positive thyroid tumors (69,70). The role of the NRs in mediating these and other metabolic changes warrants further investigation.

Our findings also highlight the issue of isoform specificity, particularly in relation to the retinoid receptor family in PTC and PPAR receptor family in FC. Although in vitro studies have suggested a potential therapeutic effect of retinoids and PPAR agonists in thyroid cancer, clinical trials have not, on the whole, demonstrated clinically significant benefit. Important factors influencing the response to these agents may include the differential expression of receptor isoforms as well as ligand specificity. As isoform-specific NR agonists and antagonists are increasingly available, studies involving modulation with these agents may be informative.

The results of our analysis represent the first systematic examination of NR expression in thyroid tumors. They provide a template for further mechanistic studies, as well as highlighting potential therapeutic targets, including novel targets for these malignancies.

Footnotes

Acknowledgments

This work was supported by the Victorian Cancer Agency through a project grant to C.G. and by the National Health and Medical Research Council through a Fellowship to P.J.F. G.E.O.M. is supported by the University of Queensland through a Vice Chancellor's Senior Research Fellowship. M.M. is supported by Monash University though a Monash University Postgraduate Scholarship. Prince Henry's Institute is supported by the Victorian Government's Operational Infrastructure Program. The authors also acknowledge the Victorian Cancer Biobank for assistance with tissue collection. The authors thank Zdenka Prodanovic for assistance with preparation of the tissue microarrays, Zdenka Prodanovic and Dr. Ann Drummond for assistance with immunohistochemistry, and Sue Panckridge for help with the preparation of the manuscript. The authors also thank the Nuclear Receptors in Breast Cancer Research Consortium for their help and support with these studies.

Author Disclosure Statement

No competing financial interests exist.

References

, Du

, Reitzel

, Xu

, Sturgis

. 2013. Impact of enhanced detection on the increase in thyroid cancer incidence in the US: review of incidence trends by socioeconomic status within the Surveillance, Epidemiology, and End Results registry, 1980–2008. Thyroid, 23:103–110.

Gasparre

, Bonora

, Tallini

, Romeo

. 2010. Molecular features of thyroid oncocytic tumors. Mol Cell Endocrinol, 321:67–76.

Sherman

. 2010. Cytotoxic chemotherapy for differentiated thyroid carcinoma. Clin Oncol (R Coll Radiol), 22:464–468.

Kapiteijn

, Schneider

, Morreau

, Gelderblom

, Nortier

JWR

, Smit

JWA

. 2012. New treatment modalities in advanced thyroid cancer. Ann Oncol, 23:10–18.

Di Vito

, De Santis

, Perrone

, Mari

, Giordano

, De Antoni

, Coppola

, Fadda

, Tafani

, Carpi

, Russo

. 2011. Overexpression of estrogen receptor-α in human papillary thyroid carcinomas studied by laser-capture microdissection and molecular biology. Cancer Science, 102:1921–1927.

Egawa

, Miyoshi

, Iwao

, Shiba

, Noguchi

. 2001. Quantitative analysis of estrogen receptor-alpha and -beta messenger RNA expression in normal and malignant thyroid tissues by real-time polymerase chain reaction. Oncology, 61:293–298.

Ceresini

, Morganti

, Graiani

, Saccani

, Milli

, Usberti

, Valenti

, Ceda

, Corcione

. 2006. Estrogen receptor (ER)-beta, but not ER-alpha, is present in thyroid vessels: immunohistochemical evaluations in multinodular goiter and papillary thyroid carcinoma. Thyroid, 16:1215–1220.

Inoue

, Oshimo

, Miki

, Kawano

, Monden

. 1993. Immunohistochemical study of estrogen receptors and the responsiveness to estrogen in papillary thyroid carcinoma. Cancer, 72:1364–1368.

Manole

, Schildknecht

, Gosnell

, Adams

, Derwahl

. 2001. Estrogen promotes growth of human thyroid tumor cells by different molecular mechanisms. J Clin Endocrinol Metab, 86:1072–1077.

10.

Rajoria

, Suriano

, Shanmugam

, Wilson

, Schantz

, Geliebter

, Tiwari

. 2010. Metastatic phenotype is regulated by estrogen in thyroid cells. Thyroid, 20:33–41.

11.

Kim

, Kim

, Ryu

, Hong

, Kim

, Shong

. 2009. Redifferentiation therapy with 13-cis retinoic acids in radioiodine-resistant thyroid cancer. Endocr J, 56:105–112.

12.

Cras

, Politis

, Balitrand

, Darsin-Bettinger

, Boelle

, Cassinat

, Toubert

M-E

, Chomienne

. 2012. Bexarotene via CBP/p300 induces suppression of NF-κB–dependent cell growth and invasion in thyroid cancer. Clin Cancer Res, 18:442–453.

13.

Liu

, Stokkel

, Morreau

, Pereira

, Romijn

, Smit

. 2008. Radioiodine therapy after pretreatment with bexarotene for metastases of differentiated thyroid carcinoma. Clin Endocrinol (Oxf), 68:605–609.

14.

Short

, Suovuori

, Cook

, Vivian

, Harmer

. 2004. A phase II study using retinoids as redifferentiation agents to increase iodine uptake in metastatic thyroid cancer. Clin Oncol (R Coll Radiol), 16:569–574.

15.

Simon

, Korber

, Krausch

, Segering

, Groth

, Gorges

, Grunwald

, Muller-Gartner

, Schmutzler

, Kohrle

, Roher

, Reiners

. 2002. Clinical impact of retinoids in redifferentiation therapy of advanced thyroid cancer: final results of a pilot study. Eur J Nucl Med Mol Imaging, 29:775–782.

16.

Vivaldi

, Miasaki

, Ciampi

, Agate

, Collecchi

, Capodanno

, Pinchera

, Elisei

. 2009. Re-differentiation of thyroid carcinoma cell lines treated with 5-Aza-2′-deoxycytidine and retinoic acid. Mol Cell Endocrinol, 307:142–148.

17.

Haugen

, Larson

, Pugazhenthi

, Hays

, Klopper

, Kramer

, Sharma

. 2004. Retinoic acid and retinoid X receptors are differentially expressed in thyroid cancer and thyroid carcinoma cell lines and predict response to treatment with retinoids. J Clin Endocrinol Metab, 89:272–280.

18.

Schmutzler

, Hoang-Vu

, Ruger

, Kohrle

. 2004. Human thyroid carcinoma cell lines show different retinoic acid receptor repertoires and retinoid responses. Eur J Endocrinol, 150:547–556.

19.

Zhu

, Zhao

, Willingham

, Cheng

. 2010. Thyroid hormone receptors are tumor suppressors in a mouse model of metastatic follicular thyroid carcinoma. Oncogene, 29:1909–1919.

20.

Thomas

, Williams

. 1989. The clonal origin of thyroid nodules and adenomas. Am J Pathol, 134:141–147.

21.

Ramelli

, Studer

, Bruggisser

. 1982. Pathogenesis of thyroid nodules in multinodular goiter. Am J Pathol, 109:215–223.

22.

Namba

, Matsuo

, Fagin

. 1990. Clonal composition of benign and malignant human thyroid tumors. J Clin Invest, 86:120–125.

23.

Harrer

, Broecker

, Zint

, Schatz

, Zumtobel

, Derwahl

. 1998. Thyroid nodules in recurrent multinodular goiters are predominantly polyclonal. J Endocrinol Invest, 21:380–385.

24.

Xing

. 2007. BRAF mutation in papillary thyroid cancer: pathogenic role, molecular bases, and clinical implications. Endocr Rev, 28:742–762.

25.

Jamieson

, Alexiadis

, Fuller

. 2004. Expression status and mutational analysis of the ras and B-raf genes in ovarian granulosa cell and epithelial tumors. Gynecol Oncol, 95:603–609.

26.

Alexiadis

, Eriksson

, Jamieson

, Davis

, Drummond

, Chu

, Clyne

, Muscat

, Fuller

. 2011. Nuclear receptor profiling of ovarian granulosa cell tumors. Horm Cancer, 2:157–169.

27.

Muscat

GEO

, Eriksson

, Byth

, Loi

, Graham

, Jindal

, Davis

, Clyne

, Funder

, Simpson

, Ragan

, Kuczek

, Fuller

, Tilley

, Leedman

, Clarke

. 2013. Research resource: nuclear receptors as transcriptome: discriminant and prognostic value in breast cancer. Mol Endocrinol, 27:350–365.

28.

Nuclear Receptor Signalling Atlas. Tissue-specific expression patterns of nuclear receptors. Thyroid gland: antomical Q-PCR expression data. Available at: www.nursa.org/qpcr.cfm?&dataType=tissue&list=tissue&mol=3C3.10090&tissue=4.4.0.0.0&strain=3&dataNumber=1&time=&x=mol&group=strain&detail=2&color=1 (accessed November 28, 2013 ).

29.

Chen

, O'Shea

. 2008. Th17 cells: a new fate for differentiating helper T cells. Immunol Res, 41:87–102.

30.

Giordano

, Kuick

, Thomas

, Misek

, Vinco

, Sanders

, Zhu

, Ciampi

, Roh

, Shedden

, Gauger

, Doherty

, Thompson

, Hanash

, Koenig

, Nikiforov

. 2005. Molecular classification of papillary thyroid carcinoma: distinct BRAF, RAS, and RET/PTC mutation-specific gene expression profiles discovered by DNA microarray analysis. Oncogene, 24:6646–6656.

31.

Rusinek

, Szpak-Ulczok

, Jarzab

. 2011. Gene expression profile of human thyroid cancer in relation to its mutational status. J Mol Endocrinol, 47:R91–R103.

32.

Frattini

, Ferrario

, Bressan

, Balestra

, De Cecco

, Mondellini

, Bongarzone

, Collini

, Gariboldi

, Pilotti

, Pierotti

, Greco

. 2004. Alternative mutations of BRAF, RET and NTRK1 are associated with similar but distinct gene expression patterns in papillary thyroid cancer. Oncogene, 23:7436–7440.

33.

Draghici

, Khatri

, Eklund

, Szallasi

. 2006. Reliability and reproducibility issues in DNA microarray measurements. Trends Genet, 22:101–109.

34.

Shippy

, Sendera

, Lockner

, Palaniappan

, Kaysser-Kranich

, Watts

, Alsobrook

. 2004. Performance evaluation of commercial short-oligonucleotide microarrays and the impact of noise in making cross-platform correlations. BMC Genomics, 5:61.

35.

Rochaix

, Monteil-Onteniente

, Rochette-Egly

, Caratero

, Voigt

, Jozan

. 1998. Reduced expression of retinoic acid receptor beta protein (RAR beta) in human papillary thyroid carcinoma: immunohistochemical and western blot study. Histopathology, 33:337–343.

36.

Takiyama

, Miyokawa

, Sugawara

, Kato

, Ito

, Sato

, Oikawa

, Kobayashi

, Kimura

, Tateno

. 2004. Decreased expression of retinoid X receptor isoforms in human thyroid carcinomas. J Clin Endocrinol Metabol, 89:5851–5861.

37.

Elisei

, Vivaldi

, Agate

, Ciampi

, Molinaro

, Piampiani

, Romei

, Faviana

, Basolo

, Miccoli

, Capodanno

, Collecchi

, Pacini

, Pinchera

. 2005. All-trans-retinoic acid treatment inhibits the growth of retinoic acid receptor beta messenger ribonucleic acid expressing thyroid cancer cell lines but does not reinduce the expression of thyroid-specific genes. J Clin Endocrinol Metab, 90:2403–2411.

38.

Hoftijzer

, Liu

, Morreau

, van Wezel

, Pereira

, Corssmit

EPM

, Romijn

, Smit

JWA

. 2009. Retinoic acid receptor and retinoid X receptor subtype expression for the differential diagnosis of thyroid neoplasms. Eur J Endocrinol, 160:631–638.

39.

Jarzab

, Wiench

, Fujarewicz

, Simek

, Jarzab

, Oczko-Wojciechowska

, Wloch

, Czarniecka

, Chmielik

, Lange

, Pawlaczek

, Szpak

, Gubala

, Swierniak

. 2005. Gene expression profile of papillary thyroid cancer: sources of variability and diagnostic implications. Cancer Res, 65:1587–1597.

40.

Zhang

. 2002. Vitamin A and apoptosis in prostate cancer. Endocr Relat Cancer, 9:87–102.

41.

Zhang

, Liu

, Lee

. 1996. Retinoid receptors in human lung cancer and breast cancer. Mutat Res, 350:267–277.

42.

X-C

. 2007. Tumor-suppressive activity of retinoic acid receptor-β in cancer. Cancer Lett, 253:14–24.

43.

Hoque

, Rosenbaum

, Westra

, Xing

, Ladenson

, Zeiger

, Sidransky

, Umbricht

. 2005. Quantitative assessment of promoter methylation profiles in thyroid neoplasms. J Clin Endocrinol Metab, 90:4011–4018.

44.

Brown

, Smart

, Sharma

, Brinkmeier

, Greenlee

, Camper

, Jensen

, Eckel

, Krezel

, Chambon

, Haugen

. 2000. Thyroid hormone resistance and increased metabolic rate in the RXR-gamma-deficient mouse. J Clin Invest, 106:73–79.

45.

Duez

, Staels

. 2009. Rev-erb-α: an integrator of circadian rhythms and metabolism. J Appl Physiol, 107:1972–1980.

46.

Cho

, Zhao

, Hatori

, Yu

, Barish

, Lam

, Chong

L-W

, DiTacchio

, Atkins

, Glass

, Liddle

, Auwerx

, Downes

, Panda

, Evans

. 2012. Regulation of circadian behaviour and metabolism by REV-ERB-[agr] and REV-ERB-[bgr]. Nature, 485:123–127.

47.

Fontaine

, Dubois

, Duguay

, Helledie

, Vu-Dac

, Gervois

, Soncin

, Mandrup

, Fruchart

, Fruchart-Najib

, Staels

. 2003. The orphan nuclear receptor Rev-Erbalpha is a peroxisome proliferator-activated receptor (PPAR) gamma target gene and promotes PPARgamma-induced adipocyte differentiation. J Biol Chem, 278:37672–37680.

48.

Migita

, Morser

, Kawai

. 2004. Rev-erbalpha upregulates NF-kappaB-responsive genes in vascular smooth muscle cells. FEBS Lett, 561:69–74.

49.

Kourtidis

, Jain

, Carkner

, Eifert

, Brosnan

, Conklin

. 2010. An RNA interference screen identifies metabolic regulators NR1D1 and PBP as novel survival factors for breast cancer cells with the ERBB2 signature. Cancer Res, 70:1783–1792.

50.

Han

, Zhou

, Kim

, Yan

, Lee

, Wu

, Lin

, Lu

, Liu

, Zeng

, Zhang

. 2009. A unique cytoplasmic localization of retinoic acid receptor-gamma and its regulations. J Biol Chem, 284:18503–18514.

51.

Matic

, Nakhel

, Lehnert

, Polly

, Clarke

, Robertson

. 2009. A novel approach to investigate the subcellular distribution of nuclear receptors in vivo. Nucl Recept Signal, 7:e004.

52.

Wang

, Yang

, Li

, He

. 2013. High cytoplasmic expression of the orphan nuclear receptor NR4A2 predicts poor survival in nasopharyngeal carcinoma. Asian Pac J Cancer Prev, 14:2805–2809.

53.

Inamoto

, Czerniak

, Dinney

, Kamat

. 2010. Cytoplasmic mislocalization of the orphan nuclear receptor Nurr1 is a prognostic factor in bladder cancer. Cancer, 116:340–346.

54.

Papadaki

, Mylona

, Giannopoulou

, Markaki

, Keramopoulos

, Nakopoulou

. 2005. PPARgamma expression in breast cancer: clinical value and correlation with ERbeta. Histopathology, 46:37–42.

55.

Vieira

, Marroqui

, Batista

, Caballero-Garrido

, Carneiro

, Boschero

, Nadal

, Quesada

. 2012. The clock gene Rev-erbalpha regulates pancreatic beta-cell function: modulation by leptin and high-fat diet. Endocrinology, 153:592–601.

56.

Pacifico

, Leonardi

. 2010. Role of NF-kappaB in thyroid cancer. Mol Cell Endocrinol, 321:29–35.

57.

Peters

, Shah

, Gonzalez

. 2012. The role of peroxisome proliferator-activated receptors in carcinogenesis and chemoprevention. Nat Rev Cancer, 12:181–195.

58.

Zeng

, Geng

, Tretiakova

, Yu

, Sicinski

, Kroll

. 2008. Peroxisome proliferator-activated receptor-δ induces cell proliferation by a cyclin E1–dependent mechanism and is up-regulated in thyroid tumors. Cancer Res, 68:6578–6586.

59.

Wang

, Wang

, Guo

, Ning

, Katkuri

, Wahli

, Desvergne

, Dey

, DuBois

. 2006. Crosstalk between peroxisome proliferator-activated receptor δ and VEGF stimulates cancer progression. Proc Natl Acad Sci USA, 103:19069–19074.

60.

, Han

, Wu

. 2006. A novel positive feedback loop between peroxisome proliferator-activated receptor-δ and prostaglandin E 2 signaling pathways for human cholangiocarcinoma cell growth. J Biol Chem, 281:33982–33996.

61.

Burdick

, Bility

, Girroir

, Billin

, Willson

, Gonzalez

, Peters

. 2007. Ligand activation of peroxisome proliferator-activated receptor-β/δ(PPARβ/δ) inhibits cell growth of human N/TERT-1 keratinocytes. Cell Signal, 19:1163–1171.

62.

Kim

, Bility

, Billin

, Willson

, Gonzalez

, Peters

. 2006. PPARβ/δ selectively induces differentiation and inhibits cell proliferation. Cell Death Differ, 13:53–60.

63.

Zhu

, Park

, Kaneshige

, Bhat

, Zhu

, Mariash

, McPhie

, Cheng

. 2000. The orphan nuclear receptor Ear-2 is a negative coregulator for thyroid hormone nuclear receptor function. Mol Cell Biol, 20:2604–2618.

64.

X-B

, Jiao

, Sun

, Xue

, Zhao

W-T

, Fan

, Wu

G-H

, Fang

. 2011. The orphan nuclear receptor EAR2 is overexpressed in colorectal cancer and it regulates survivability of colon cancer cells. Cancer Lett, 309:137–144.

65.

Ichim

, Atkins

, Iscove

, Wells

. 2011. Identification of a role for the nuclear receptor EAR-2 in the maintenance of clonogenic status within the leukemia cell hierarchy. Leukemia, 25:1687–1696.

66.

Máximo

, Lima

, Prazeres

, Soares

, Sobrinho-Simões

. 2012. The biology and the genetics of Hürthle cell tumors of the thyroid. Endocr Relat Cancer, 19:R134–R147.

67.

Ganly

, Ricarte Filho

, Eng

, Ghossein

, Morris

LGT

, Liang

, Socci

, Kannan

, Mo

, Fagin

, Chan

. 2013. Genomic dissection of Hurthle cell carcinoma reveals a unique class of thyroid malignancy. J Clin Endocrinol Metab, 98:E962–E972.

68.

Goffredo

, Roman

, Sosa

. 2013. Hurthle cell carcinoma: a population-level analysis of 3311 patients. Cancer, 119:504–511.

69.

Grabellus

, Worm

, Schmid

, Sheu

. 2012. The BRAF V600E mutation in papillary thyroid carcinoma is associated with glucose transporter 1 overexpression. Thyroid, 22:377–382.

70.

Feng

, Gao

, Wang

, Yu

, Zhang

, Guan

, Shan

, Teng

. 2013. Aberrant overexpression of pyruvate kinase M2 is associated with aggressive tumor features and the BRAF mutation in papillary thyroid cancer. J Clin Endocrinol Metab, 98:E1524–E1533.