Oncogenic Actions of the Nuclear Receptor Corepressor (NCOR1) in a Mouse Model of Thyroid Cancer

Studies have suggested that the nuclear receptor corepressor 1 (NCOR1) could play an important role in human cancers. However, the detailed molecular mechanisms by which it functions in vivo to affect cancer progression are not clear. The present study elucidated the in vivo actions of NCOR1 in carcinogenesis using a mouse model (ThrbPV/PV mice) that spontaneously develops thyroid cancer. ThrbPV/PV mice harbor a dominantly negative thyroid hormone receptor β (TRβ) mutant (denoted as PV). We adopted the loss-of-the function approach by crossing ThrbPV mice with mice that globally express an NCOR1 mutant protein (NCOR1ΔID) in which the receptor interaction domains have been modified so that it cannot interact with the TRβ, or PV, in mice. Remarkably, expression of NCOR1ΔID protein reduced thyroid tumor growth, markedly delayed tumor progression, and prolonged survival of ThrbPV/PVNcor1 ΔID/ΔID mice. Tumor cell proliferation was inhibited by increased expression of cyclin-dependent kinase inhibitor 1 (p21waf1/cip1; Cdkn1A), and apoptosis was activated by elevated expression of pro-apoptotic BCL-Associated X (Bax). Further analyses showed that p53 was recruited to the p53-binding site on the proximal promoter of the Cdkn1A and the Bax gene as a co-repressor complex with PV/NCOR1/histone deacetylas-3 (HDAC-3), leading to repression of the Cdkn1A as well as the Bax gene in thyroids of ThrbPV/PV mice. In thyroids of ThrbPV/PVNcor1 ΔID/ΔID mice, the p53/PV complex could not recruit NCOR1ΔID and HDAC-3, leading to de-repression of both genes to inhibit cancer progression. The present studies provided direct evidence in vivo that NCOR1 could function as an oncogene via transcription regulation in a mouse model of thyroid cancer.


Introduction
Thyroid hormone receptors (TRs) are critical in mediating the genomic actions of the thyroid hormone (T3) in growth, development, differentiation, and maintaining metabolic homeostasis. Two TR genes, a and b, located on two different chromosomes, encode three major T3 binding TR isoforms (a1, b1, and b2). Studies using genetically engineered mice showed that, in vivo, TR isoforms have common functions, but can also exert isoform-dependent actions in target tissues [1,2]. The transcription activity of TR is regulated at multiple levels. In addition to the regulation by T3 and types of DNA binding elements in the promoters of target genes, the transcriptional activity of TR is fine-tuned by a host of nuclear coactivators and corepressors [3,4]. In the absence of T3, the unliganded TR recruits the nuclear receptor corepressor 1 (NCOR1) and the nuclear receptor corepressor 2/silencing mediator for retinoid and thyroid hormone receptors (NCOR2/SMRT) for transcriptional repression. Binding of T3 leads to a conformational change in the TR that releases the NCOR1/NCOR2 complex and allows for the recruitment of a multiprotein coactivator complex for transcriptional activation [5].
The important regulatory role of NCOR1 in receptor actions is evident in its aberrant interaction with receptors underlying human diseases such as resistance to thyroid hormone (RTH) [6,7] and lipodystrophic severe insulin resistance [8]. RTH is caused by mutations in the THRB gene [9,10]. Mutations of the peroxisome proliferator-activated receptor c (PPARc) lead to lipodystrophic severe insulin resistance [8]. In addition, aberrant interaction of NCOR1/SMRT with the fused gene products of the retinoic acid receptor a gene (RARa) or with the promyelocytic leukemia gene (PML) (PML-RAR-a) or with the promyelocytic leukemia zinc finger gene (PLZF) (PLZF-RAR-a) results in blocking myeloid differentiation [11]. The involvement of NCOR1 in other human cancers such as breast and bladder cancers was demonstrated by a close association of NCOR1 abnormal expression with cancer development [12,13,14,15]. Recent studies by The Cancer Genome Atlas (TCGA) Research Network also uncovered homozygous gene deletion of NCOR1 in some patients with colon and rectum adenocarcinoma [16], prostate adenocarcinoma [17], ovarian carcinoma [18] or liver hepatocellular carcinoma (unpublished-Provisional in TCGA data base). Moreover, nonsense mutations and splicing variants of NCOR1 were also found in patients with breast cancer [19]. While these association studies raised the possibility that NCOR1 could act to affect cancer progression, direct evidence of its roles in carcinogenesis in vivo is still lacking.
The availability of a mouse model that spontaneously develops metastatic follicular thyroid cancer (FTC) provided us with a powerful tool to assess the role of NCOR1 in cancer development and progression. The Thrb PV/PV mouse harbors a knockin dominant negative mutation, known as PV, in the Thrb gene locus [20]. The PV mutation was identified in a patient suffering from resistance to thyroid hormone [21]. As Thrb PV/PV mice age, their thyroids undergo pathological changes from hyperplasia to capsular and vascular invasion, anaplasia, and eventual metastasis to the lung [22]. The pathological progression, route, and frequency of metastasis in Thrb PV/PV mice are similar to that in human FTC. Extensive molecular analyses of altered signaling pathways show that, as found in human FTC, Thrb PV/PV mice exhibit aberrant signaling pathways that include constitutive activation of phosphatidyl inositol 3-kinase (PI3K)/Akt [23,24] and integrin-Src-MAPK signaling [25] and aberrant accumulation of the oncogenic pituitary tumor transforming gene protein (PTTG) [26,27] and ß-catenin [28]. Thus, the Thrb PV/PV mouse model faithfully recapitulates the molecular aberrations found in human thyroid cancer and is indeed a preclinical mouse model of FTC.
In the present studies, we adopted the loss-of-the function approach by crossing Thrb PV with mice that globally express an NCOR1 mutant protein (NCOR1DID). In this NCOR1DID mutant protein, two most amino terminal receptor interaction domains termed RID 3 and RID2, are missing. This mutant cannot associate with TR, as RID3 is absolutely required for NCOR1-TR interaction [29,30,31]. Consistently, we also showed that NCOR1DID mutant protein cannot interact with TRbPV in vivo [32]. The lack of interaction of TRbPV with NCOR1DID results in the amelioration of symptoms of resistance to thyroid hormone in Thrb PV Ncor1 DID/DID mice, indicative that NCOR1 regulates the dominant negative action of TRb mutants in vivo [32].
The present studies show that the expression of NCOR1DID in the thyroid of Thrb PV/PV mice inhibited tumor growth, prolonged survival, and delayed cancer progression. Tumor cell proliferation was inhibited by the increased expression of cyclin-dependent kinase inhibitor 1 (p21/WAF1; Cdkn1A), and the apoptosis was induced by the activated expression of the Bcl-2-associated X protein (BAX). Both the Cdkn1A and the BAX genes are direct downstream target genes of the tumor suppressor, p53. Detailed molecular analyses showed that the lack of receptor interaction domain in NCOR1DID led to an inability of the p53/PV complex to recruit the NCOR1/histone deacetylase-3 repressor complex to the promoters of these target genes, resulting in increased repression of these genes to inhibit cancer progression. The present study provided direct evidence in vivo to demonstrate that the NCOR1 could function as an oncogene via transcription regulation in thyroid carcinogenesis in a mouse model.

Mouse Strains
The animal study was carried out according to the protocol approved by the National Cancer Institute Animal Care and Use Committee. Mice harboring the Thrb PV gene (Thrb PV mice) were prepared and genotyped as described earlier [20]. Thrb PV/ PV Ncor1 DID/DID mice were bred by first crossing Thrb +/+ Ncor1 DID/ + mice [29] with Thrb PV/+ mice and then by crossing Thrb PV/ + Ncor1 DID/DID mice with Thrb PV/+ Ncor1 DID/DID mice [32]. Mice with different genotypes used in the present study were intercrossed several generations, and littermates with a similar genetic background were used in all experiments. Mice were monitored until they became moribund and therefore euthanized. Thyroids and other tissues were collected from Thrb PV/PV Ncor1 +/+ mice and Thrb PVPV Ncor1 DID/DID mice for weighing, histological analysis, and molecular and biochemical studies.

Histological Analysis
Thyroid gland, lung, and heart were dissected, fixed in 10% neutral buffered formalin (Sigma-Aldrich, St. Louis, MO), and subsequently embedded in paraffin. Sections of 5-mm thickness were prepared and stained with hematoxylin and eosin (H&E). For each animal, single random sections through the thyroid, through the lung, and through the heart were examined.

Immunohistochemistry
IHC was performed as previously described with some modification [25]. Formalin-fixed paraffin thyroid sections were deparaffinized, rehydrated, heated to 98uC in 0.05% citraconic anhydride, pH 7.4 (Sigma-Aldrich, St. Louis, MO), for 1 hour and then blocked for 1 hour in 2% normal goat serum at room temperature. After washing in phosphate-buffered saline, slides were incubated at 4uC overnight with Ki-67 primary antibody (1:300 dilution; Thermo Scientific, Fremont, CA; #RB-9043-P0). After washing, slides were incubated with goat anti-rabbit secondary antibody for 1 hour at room temperature and rinsed in phosphate-buffered saline. Slides were then incubated in 3,39diaminobenzidene (DAB substrate kit for peroxidase; Vector Laboratories, Burlingame, CA; SK-4100), and after staining development, they were counterstained in Gill's Hematoxylin, rinsed and mounted in Permount (Fisher Scientific, Pittsburgh, PA). The proliferative index was calculated as the percentage of Ki-67-positive nuclei to the total number of nuclei on the thyroid section. Counting was performed using National Institutes of Health (NIH) IMAGE software (ImageJ 1.34 s; Wayne Rasband, NIH, Bethesda, MD).

Statistical Analysis
All data are expressed as mean 6 the SEM. Differences between groups were examined for statistical significance using Student's t test with the use of GraphPad Prism 4.0a (GraphPad Software); p,0.05 is considered statistically significant.

NCOR1DID Prolongs Survival and Delays Thyroid Carcinogenesis of Thrb PV/PV Mice
We have previously shown that Thrb PV/PV mice have high mortality caused by FTC [22]. We monitored the effects of the expression of NCOR1DID on thyroid carcinogenesis by first comparing the survival of Thrb PV/PV Ncor1 DID/DID mice with that of Thrb PV/PV Ncor1 +/+ mice. Mice were monitored until they became moribund with signs of palpable tumors, rapid weight loss, hunched postures, and labored breathing. Figure 1A shows that Thrb PV/PV Ncor1 DID/DID mice survived significantly longer (p,0.01; 50% survival age: 11.3 months, n = 29) than did Thrb PV/PV Ncor1 +/+ mice (50% survival age: 9.3 months, n = 58) during the 15-month observation period. Figure 1B shows that the expression of NCOR1DID led to a significant 35% reduction in thyroid weight in Thrb PV/PV Ncor1 DID/DID mice (data set 2 vs. 1; p,0.0001).
The effect of NCOR1DID on the pathological progression was assessed by histopathological analysis as the Thrb PV/PV Ncor1 DID/DID mice aged. Figure 2A shows representative pathological features of the thyroids and lungs from the 3-to 15-month-old Thrb PV/ PV Ncor1 +/+ and Thrb PV/PV Ncor1 DID/DID mice (left and right columns, respectively). In the thyroids of Thrb PV/PV Ncor1 +/+ mice at the age of 7 months, advanced vascular invasion and focal anaplasia were frequently observed (arrows in Figure 2Aa and 2Ac, respectively). Moreover, lung metastases (Figure 2Ae, arrows) were frequent in Thrb PV/PV Ncor1 +/+ mice. In Thrb PV/PV Ncor1 DID/ DID mice, however, vascular invasion was rarely observed (Figure 2Ab), but with no detectable anaplasia ( Figure 2Ad) and a marked reduction in the occurrence of lung metastasis (Figure 2Af). These pathohistological observations are summarized in Figure 2B. At the younger age of 3-5 months, a lower occurrence of capsular invasion (,20%) was observed for Thrb PV/ PV Ncor1 DID/DID mice than for Thrb PV/PV Ncor1 +/+ mice (Figure 2Ba), but a similar occurrence frequency of capsular invasion was detected for both mutant mice at older age (.7 months of age). No vascular invasion, anaplasia, or lung metastases were found in Thrb PV/PV Ncor1 DID/DID mice (Figure 2Bb, c, and d) at younger ages of 3-5 months. For mice older than 7 months, the occurrence frequency of vascular invasion, anaplasia, and lung metastasis were 80%, 15%, and 60%, respectively, for Thrb PV/PV Ncor1 +/+ mice (Figure 2Bb, c, and d). However, the occurrence of vascular invasion and lung metastasis were 14% and 10%, respectively, for Thrb PV/PV Ncor1 DID/DID mice, with no occurrence of anaplasia. Taken together, these results indicate that the expression of NCOR1DID delayed thyroid cancer progression and blocked loss of differentiation (anaplasia) in Thrb PV/PV mice. The fact that thyroid growth was reduced in Thrb PV/PV Ncor1 DID/ DID mice ( Figure 1B) prompted us to ask whether thyroid tumor cell proliferation was inhibited. We therefore examined the protein abundance of the nuclear proliferation marker, Ki-67, by immunohistochemical analysis. Figure 3A.I-a and -b show representative examples of intensive nuclear staining in thyroids of two Thrb PV/PV Ncor1 +/+ mice. In contrast, fewer nuclei were immuno-stained with Ki-67 proliferation marker in thyroids of Thrb PV/PV Ncor1 DID/DID mice ( Figure 3A.I-c and -d, n = 2). Cells with Ki-67 immuno-stained nuclei were counted and the quantitative data are shown in Figure 3A.II. The quantitative analysis shows that the number of thyroid cells with Ki-67 stained nuclei was 50% lower in Thrb PV/PV Ncor1 DID/DID mice than in Thrb PV/PV Ncor1 +/+ mice, indicating decreased cell proliferation in the thyroid of Thrb PV/PV Ncor1 DID/DID mice.
This finding was further supported by the biochemical analysis in which the protein abundance of a key cell cycle regulator, cyclin D1, was markedly lower in the thyroids of Thrb PV/PV Ncor1 DID/DID mice ( Figure 3B.I, panel a, lanes 3 & 4) than in Thrb PV/PV Ncor1 +/+ mice (lanes 1 & 2). Moreover, the reduced cyclin D1 led to a lower protein abundance of phosphorylated retinoblastoma protein (pRb) ( Figure 3B.I, panel b, compare lanes 1 & 2 with lanes 3 & 4) without changing the total Rb protein levels (panel c). The reduction in the phosphorylated Rb impeded the progression of the cell cycle from the G1 to the S phase. Consistently, the protein abundance of cyclin-dependent kinase inhibitor p21 and p27 ( Figure 3B.I, panels d and e, respectively) was also lower in the thyroids of Thrb PV/PV Ncor1 DID/DID mice than in Thrb PV/PV Ncor1 +/+ mice (compare lanes 3-4 to lanes 1-2). The quantitative data of the band intensities are shown in Figure 3B.II. Collectively, these results indicate that the expression of the NCOR1DID led to inhibition of tumor cell proliferation, in part, by delaying the G1-S cell cycle progression.
Whether apoptosis also contributed to decreased thyroid tumor growth shown in Figure 1B was also evaluated by examining the key regulators of apoptosis. The protein abundance of Bcl-2associated X protein (BAX), which promotes apoptosis, was higher in thyroids of Thrb PV/PV Ncor1 DID/DID mice than in Thrb PV/ PV Ncor1 +/+ mice (Figure 4Aa, compare lanes 3 & 4 with 1 & 2). PUMA, which is a BH3 (Bcl-2 homology domain 3)-only protein that induces apoptosis through the mitochondria pathway, was also higher in thyroids of Thrb PV/PV Ncor1 DID/DID mice than in Thrb PV/PV Ncor1 +/+ mice (Figure 4Ab These results indicate that in addition to delaying the G1-S cell cycle progression, the increased apoptotic activity in thyroids of Thrb PV/PV Ncor1 DID/DID mice also contributed to lower thyroid growth.   The findings that protein abundances of p21 and BAX were altered as shown above prompted us to examine whether their expression at the mRNA level was also affected. Indeed, we found that Cdkn1A (p21 WAF1 ) mRNA levels were significantly higher in the thyroid of Thrb PV/PV Ncor1 DID/DID mice than in Thrb PV/ PV Ncor1 +/+ mice ( Figure 5A, bar 2 vs. bar 1). Similarly, Bax mRNA level was higher in the thyroid of Thrb PV/PV Ncor1 DID/DID mice than in Thrb PV/PV Ncor1 +/+ mice ( Figure 5B, bar 4 vs. bar 3).
These two genes are directly regulated by p53 [35,36]. We therefore hypothesized that altering the activity of p53 in thyroids of Thrb PV/PV Ncor1 DID/DID mice would lead to activation of transcription of the Cdkn1A and Bax genes. We first examined whether the expression of p53 mRNA and protein was altered, thereby changing the activity in the thyroid of Thrb PV/PV Ncor1 DID/ DID mice. We found that the expression of NCOR1DID had no effects of the expression of p53 at the mRNA level (bars 5 and 6, Figure 5A). In addition, we found that the protein abundance of p53 was similarly low in the thyroid of Thrb PV/PV Ncor1 +/+ and Thrb PV/PV Ncor1 DID/DID mice ( Figure 5B, upper panel, lanes 3 and  4, respectively). Therefore, the higher Cdkn1A and Bax mRNA levels were not due to activation of p53 transcription activity by the elevated p53 protein expression. Because it has been previously shown that wild-type TRs physically interact with p53 and negatively regulate the transcriptional activity of p53 [37,38], we next explored the possibility that the activity of p53 could be altered via interaction with PV in thyroids of mutant mice. Coimmunoprecipitation analysis showed that PV was associated with p53 in the thyroid of Thrb PV/PV Ncor1 +/+ mice ( Figure 5B, upper panel, lane 1). PV was also similarly associated with p53 in the thyroid of Thrb PV/PV Ncor1 DID/DID mice ( Figure 5B, upper panel, lane 2). Importantly, co-immunoprecipitation further showed that PV was also associated with NCOR1 in the thyroid nuclear extracts of Thrb PV/PV Ncor1 +/+ mice ( Figure 5B, lower panel, lane 1), but did not interact with NCOR1DID in the thyroid nuclear extracts of Thrb PV/PV Ncor1 DID/DID mice (lane 2). Lanes 3 and 4 show that a similar amount of NCOR1 and NCOR1DID, respectively, was detected by direct Western blot analysis ( Figure 5B, lower panel). These data indicate that PV interacted with p53 and NCOR1 in a ternary complex (p53/PV/NCOR1) in the thyroid of Thrb PV/PV Ncor1 +/+ mice, but formed only a p53/PV complex in the thyroid of Thrb PV/PV Ncor1 DID/DID mice.
The findings that Cdkn1A and Bax mRNA were higher in the thyroid of Thrb PV/PV Ncor1 DID/DID than Thrb PV/PV Ncor1 +/+ mice ( Figure 5A) provided us with a tool to further investigate the mechanisms by which NCOR1 regulated thyroid carcinogenesis. We therefore used a chromatin immunoprecipitation (ChIP) assay to understand how the transcriptional activity of p53 was activated in the thyroid of Thrb PV/PV Ncor1 DID/DID mice to drive the expression of the Cdkn1A and Bax genes to inhibit tumor growth. Figure 5C shows that when anti-p53 antibodies were used in the ChIP assays, a strong signal above the background (bar 2 vs. bar 1) was detected to indicate that p53 was recruited to the p53 binding sites on the promoter region of the Cdkn1A gene (-1961 to -1942 bp; see Figure 6A) [35] in the thyroid of Thrb PV/PV Ncor1 +/+ mice ( Figure 5C, bar 2). Importantly, when anti-PV antibodies were used in the ChIP, significant recruitment of PV to the promoter of the Cdkn1A gene was detected ( Figure 5C, bar 3). Concurrently, using anti-NCOR1 and histone deacetylase-3 (HDAC-3) antibodies, we also found that NCOR1 and HDAC-3 were recruited to the p53/PV complexes ( Figure 5C, bars 4 and 5, respectively). These results indicate that in thyroid tumors of Thrb PV/PV Ncor1 +/+ mice, the p53/PV/NCOR1/HDAC-3 repressor complex was recruited to the promoter of the Cdkn1A gene to repress the expression of the Cdkn1A gene. In contrast, in thyroids of Thrb PV/PV Ncor1 DID/DID mice ( Figure 5C), although p53 (bar 7) and PV (bar 8) were recruited to the p53 binding sites on the promoter of the Cdkn1A gene, no recruitment of NCOR1DID or HDAC-3 was found with the p53/PV complexes ( Figure 5C, bars 9 and 10, respectively; see also Figure 6B). Similar ChIP analysis shows the significant recruitment of p53/PV/NCOR1/HDAC-3 to the p53 binding sites (nt -781 to -768) on the promoter of the Bax gene in the thyroids of Thrb PV/PV Ncor1 +/+ mice ( Figure 5D; bars 2-5). But in the thyroids of Thrb PV/PV Ncor1 DID/DID mice, only significant recruitment of p53/PV to the p53 binding sites on the Bax promoter was detected ( Figure 5D, bars 7 and 8), but no recruitment of NCOR1DID or HDAC-3 to the promoter of the Bax promoter was observed (bars 9 and 10 vs. bar 6). These in vivo results indicate that in thyroids of Thrb PV/PV Ncor1 DID/DID mice, the lack of interaction of PV with NCOR1DID reversed the repressive effect of NCOR1/HDAC-3. This, in turn, re-activated the p53 transcription activity to increase the expression of the Cdkn1A and Bax genes, thus delaying tumor progression by inhibiting tumor cell proliferation and activation of apoptosis.

Discussion
Using a preclinical mouse model of thyroid cancer, we showed that the expression of NCOR1DID in the thyroid of Thrb PV/PV mice decreased tumor growth, delayed cancer progression, and increased survival. Analyses of key regulators underlying these reduced cancer phenotypes identified the increased expression of two p53 downstream direct target genes, the Cdkn1A and Bax genes. We further showed that p53 was associated with PV on the p53 binding sites of the promoters of the Cdkn1A and the Bax genes. However, NCOR1DID, lacking the receptor interaction domain, cannot be recruited by p53/PV to form the NCOR1-DID/HDAC-3 associated repressor complex, leading to the activation of these two genes. The increased expression of these genes resulted in the decreased proliferation and increased apoptosis of tumor cells. These findings provided direct evidence to demonstrate that, in vivo, NCOR1 can function as an oncogene to drive thyroid cancer progression. One mechanism we uncovered is via transcription regulation by constitute association with p53/PV to recruit the NCOR1/HDAC associated repressor complex, which leads to aberrant chromatin acetylation around the promoters of the p53 target genes (see Figure 6A). The repression of p53 target genes that function as tumor suppressors is therefore attenuated, and leads to cancer progression. Such oncogenic roles of NCOR1 have been demonstrated in acute promyelocyte leukemia. A fusion between the retinoic acid receptor a (RARa) and either promyelocyte leukemic (PML) or promyelocyte leukemic zinc finger (PLZF) genes sustain NCOR1 interaction. Consequently, RARa-mediated cell differentiation is blocked because of a condensed chromatin structure around the promoters of the RARa target genes [11]. However, at present we cannot exclude the possibility that NCOR1 could act via other pathways in addition to p53 signaling. Such possibilities await future exploration.
The present studies also uncovered a critical role of NCOR1 in regulating the oncogenic actions of a TR mutant, PV, via the p53 network of signaling. Previously, we showed that PV physically interacts with the C-terminal SH domain of the p85 subunit of phosphatidylinositol 3-kinase (PI3K), resulting in the activation of PI3K-AKT signaling to promote thyroid carcinogenesis of Thrb PV/ PV mice [23]. Subsequently, we discovered that NCOR1, when present in the cytoplasm, could compete with PV for binding to the same C-terminal SH domain of the p85 subunit, thereby attenuating the aberrant activation of PI3K-AKT signaling by PV [39]. The present studies showed that recruitment of nuclear NCOR1/HDAC-3 by DNA-bound p53/PV attenuated the expression of two p53 target genes, Cdkn1A and Bax, to promote tumor cell proliferation and to decrease apoptosis, respectively. Therefore, NCOR1 acts in dual modes on the oncogenic actions of PV via nuclear and extranuclear actions to affect thyroid carcinogenesis of Thrb PV/PV mice. However, it is reasonable to postulate that the nuclear action could be the predominant mode of action as NCOR1 is mainly located in the nucleus. Moreover, p53, as a central node of a complex network of signaling, when bound to DNA and associated with PV/NCOR1/HDAC-3 repressor complex, could broadly attenuate the expression of an array of tumor suppressors to promote carcinogenesis. This mode of action is exemplified by the regulation of the expression of the Cdkn1A and the Bax genes (see Figure 6). Identification of additional p53 target genes affected by the recruitment of the p53/PV/NCOR1/HDAC-3 repressor complex to the p53 binding sites would reveal the breadth of the p53 network extensively affecting thyroid carcinogenesis. However, such studies are for future investigation.
Moreover, we also cannot exclude the possibility that NCOR1-DID could act via other pathways other than p53 signaling. NCOR1 and the homolog NCOR2/SMRT are two intensively studied co-repressors [40,41]. They both contain a conserved bipartite nuclear-receptor-interaction domain (RID) and three independent repressor domains [42]. These corepressors interact with nuclear receptors via the motif, termed the CoRNR box [43]. However, numerous biochemical studies have suggested that RAR preferentially recruits SMRT, whereas TR preferentially recruits NCOR1 [44,45], mainly because of specific sequences on the motifs as well as a TR-specific interaction domain present in NCOR1, but not SMRT [44]. Accordingly, in the present study, we used the mutant Ncor1 DID mice in which the TR-interaction domains were deleted, thus preventing interaction with TR or PV [30]. We also focused our molecular analysis on the recruitment of the PV/NCOR1/HDAC-3 repressor complex by p53 to the p53 binding sites of target genes ( Figure 5). However, at present, we cannot exclude the possibility that p53/PV could also form a PV/ SMRT/HDAC-3 repressor complex, thereby contributing to thyroid carcinogenesis of Thrb PV/PV mice by repression of tumor suppressors that are downstream targets of the p53-signaling. Currently, the potential oncogenic actions of SMRT have not been extensively investigated. Whether SMRT plays a role in thyroid carcinogenesis would be interesting to address in future studies.
The present results demonstrated the amelioration of the cancer phenotypes by the expression of NCOR1DID in the thyroid of Thrb PV/PV mice. However, the expression of NCOR1DID cannot completely block the development of thyroid cancer. These findings suggest that while activation of the p53-downstream tumor suppressors (e.g., the Cdkn1A and the Bax genes) by NCOR1DID is crucial, other p53-independent pathways driven by the oncogenic actions of PV are ongoing to continuously propel thyroid cancer progression. Previously, we found that PV, via extranuclear actions, aberrantly activates PI3K/AKT [23,24] and integrin-Src-MAPK signaling [25] and abnormally accumulates the oncogenic pituitary tumor transforming gene protein (PTTG) [26,27] and ß-catenin [28]. These preclinical findings clearly show that there are multiple oncogenic molecular pathways interconnected to drive the progression. Thus, it is important to consider using combined-drug modalities to concurrently target different pathways for effective treatment. We have previously shown that treatment of Thrb PV/PV mice with a PI3K inhibitor, LY294002, effectively delays thyroid tumor progression and inhibits distant metastasis [46]. But inhibition of the PI3K/AKT pathway alone is not sufficient to completely block cancer development and progression. The present studies suggest that inhibitors of histone deacetylases could be used together in conjunction with inhibitors of PI3K/AKT or integrin-Src-MAPK pathways. HDAC inhibitors have been shown to inhibit cell proliferation [47,48,49,50] and in combination with all-trans-retinoic acid to increase redifferentiation of human thyroid cancer cell lines [51]. Completed Phase I and II clinical trials using various HDAC inhibitors have shown promise for treating radioiodine-refractory metastatic thyroid cancer [52]. Thus, treatment combining HDAC inhibitors with newer generations of inhibitors in the PI3K/AKT and integrin-Src-MAPK pathways could improve the efficacy offered by a single-drug treatment.